GOVERNMENTS of most developing countries seek to encourage the establishment and growth of financial intermediaries (defined as all financial institutions except commercial banks) by all kinds of legislative and financial aid. The ultimate success and independent development of financial intermediaries depends, however, on far more than just government assistance. The economic environment must be suitable to long-term financial activity in that income earners must be willing to save in monetary form over long periods, potential borrowers must be willing and able to invest in long-term productive (as opposed to speculative) activity; the rate of interest must be positive and reflect the cost of capital in that country, and the sectors in which the intermediaries are expected to specialize must not be throttled by price controls or unwarranted neglect. In most economies these conditions can be taken for granted; in a few, those that suffer from considerable inflation, they cannot.
Inflation damages all the basic conditions of success for financial intermediaries. It reduces savings as a whole by introducing uncertainty about future prices and increasing the burden of consumption costs for people whose income declines in real terms, and it substitutes nonmonetary savings, which are useless for productive investment, for monetary savings, which can be productively employed. It enhances the attractiveness of short-term speculative investment while decreasing that of long-term productive investment. It damages the essential allocative function of interest by lowering or nullifying it in real terms, so that credit is rationed haphazardly and with little relation to its true value. And often it causes the government to impose price controls on selected sectors of the economy, especially agriculture and public utilities, in order to control inflation, thus leading to a reduction in activity in these sectors. All this results in the uneconomic allocation of whatever short-term monetary savings take place, with an unwarranted redistribution of value in favor of the borrowers, while long-term monetary transactions are severely constricted, both because the value of money declines and because the rate of decline is itself uncertain.
It is clear that financial intermediaries would find it difficult to thrive under these conditions if not permitted to protect themselves by raising interest rates or by other measures. The best remedy would of course be to halt inflation, and as difficult as this task is, countries do succeed in it, even if the success can never be regarded as final. Countries suffering from the severer forms of inflation, however, will not be able to stop it suddenly without suffering major dislocations in their economies, and may only be able to tackle it slowly over a period of years. And some countries may experience a new bout of inflation or a resurgence of partly controlled inflation. If inflation is accepted for these countries as a fact of economic life for some time to come, the best method of counteracting its ill effects would be to reproduce somehow the conditions that would obtain under stability—that would, in other words, enable the intermediaries to attract long-term savings by offering stable value as well as a positive rate of interest in terms of real value (the purchasing power of money), and to lend for long periods with stable real value at an appropriate real rate of interest so that capital is used where most needed for productive purposes. Such a method would enable long-term financing to take place in the sectors selected without most of the deleterious effects arising from inflation.
Various techniques for counteracting inflation may be suggested which more or less approach the ideal. They can be classified broadly into two types: techniques that adjust the financial transactions to the actual rate of inflation and techniques that do not. Let us start with the latter.
Techniques Not Adjusting to Inflation
The natural reaction of financial intermediaries when faced with the reduced supply of, and increased demand for, credit caused by inflation would be to raise borrowing and lending rates until both were brought in line again. If frustrated in this reaction by government controls, they would try to institute higher rates by hidden charges and subsidies. Since, however, terms of lending are generally fixed at the outset of the transaction and since the future course of price increases is uncertain (and usually very erratic), it would be almost impossible to ensure that the higher rates of interest set or agreed upon by both parties would in fact only compensate for the decline in value of money and not favor one party over another. If, for instance, inflation were expected by everyone to range at 15 per cent for the next year and interest rates were raised by a corresponding amount to compensate for it, and in fact prices rose in general by only 5 per cent, the lenders would gain a windfall profit (while borrowers would lose by a corresponding amount). Furthermore, the longer the period of the loan the greater the risk that the initial interest rate adjustment for inflation would be wrong, and that one or the other party would lose. Clearly, therefore, it would be to the advantage of both lenders and borrowers to shorten the life of the loan to a period for which expectations of inflation were held with some degree of confidence and there were some uniformity in their expectations. Where inflation is erratic, this period may be very short, perhaps only a few months; it may rise to a year or more where the inflation is fairly low and does not fluctuate too much.
In either event, the compensation offered by raising interest rates prior to the loan, and thus not based ex post on the actual inflation, fails to provide a satisfactory basis for long-term credit in countries where inflation is quite high; consequently, it becomes very difficult to finance long-term investment by borrowing, since loans would have to be repaid and renegotiated at the end of each period for which the transactions were undertaken. In each transaction one party may lose to the other owing to the vagaries of inflation, and render the arrangements for refinancing all the more difficult. It is hard to envisage a market for long-term credit developing in such circumstances. Intermediary institutions would find it especially difficult to operate because they would have to satisfy both borrowers and lenders with right rates of interest over long periods, and obviously they would make as many errors in forecasting as anyone else.
There is also an institutional problem here. Many central banks feel obliged to regulate the levels of interest rates charged by financial institutions in their countries and to impose ceilings on them regardless of the rate at which money is losing its real value. In Brazil, for instance, the National Development Bank was allowed to charge 14 per cent per annum above the lending rate as compensation for inflation after August 1965, a period when inflation ranged much higher. Perhaps the charging of interest rates of, say, 80 per cent per annum appears usurious to the central bank concerned, though it may be barely appropriate if the inflation is expected to be 75 per cent in the same period. Or perhaps the government, optimistic by necessity, hopes to reduce the rate of price increase in the coming year and does not wish to sanction officially such a high rate of inflation. The imposition of a ceiling on interest charges has one of two bad effects—it imposes a loss, sometimes substantial, on the lender, and thus chokes off the supply of savings, or it causes lenders to resort to devious hidden charges to compensate for the loss of value, rendering financial operations more difficult and more risky.
In sum, therefore, techniques of this sort do not provide a satisfactory remedy to the ills caused by high inflation. The inherently erratic and unpredictable nature of inflation prevents any compensation not based on the actual rate of price rise from being absolutely reliable and unbiased for long periods and thus from serving as a basis for long-term financial agreements which are largely matters of trust and fair dealing. Where inflation is quite low, however, high interest rates may be acceptable to all parties as an adequate solution, since the risk of loss is not very large and fixing the terms of the loan makes financial calculations much easier.
Techniques that Adjust Loans to Inflation
There is a variety of devices available which provide that long-term financial contracts maintain their real value, all based on different means of adjusting the nominal value of the contracts to their real value in terms of the purchasing power of money. Such devices, known as “purchasing power guarantees” or “value linking” of financial obligations, are by no means newcomers to the financial scene; on the contrary, they have been advocated by classical economists such as Marshall and Jevons, and later by Keynes, and in recent times have been proposed by various economists and used by many countries, including Austria, Finland, France, and Israel.
In spite of their wide application, orthodox financial circles have resisted value linking devices because they introduce a dual standard of value, with money being used for current transactions and some index of inflation for deferred obligations. Experience has not shown, however, that this dual standard has any special disturbing effects on the financial system. In the countries in which it has been introduced, it has not led to a loss of confidence in the currency or to a loss of faith in the authorities’ ability to control inflation, though it has been argued with some reason that a strict anti-inflationary policy should somehow break the expectations of rising prices and consequently abolish value linking. Where inflation is a given fact for some time to come, however, value linking may offer substantial benefits.
How do value linking devices work? A loan is made by a financial intermediary for, say, ten years, on condition that the value of the loan is maintained in real terms. An index of prices is chosen, such as the cost of living index, and at specified intervals the repayable sum is increased in monetary terms by the amount of rise in the index. The increase may be applied only to the interest component of repayment, or to the principal, or to both. When the final repayment has been made in the tenth year, the total sum is equal in constant value to the sum lent initially plus the relevant rate of interest charged. Similarly, the intermediary institution guarantees the value of savings deposited with it and pays interest on the constant real value of the deposit. The total effect is equivalent to a situation where no inflation exists and neither borrower nor lender gains unduly at the other’s expense.
Problems in Value Linking
There are many difficulties and complications in the introduction of value linking devices, and these can be grouped under three broad headings. The first is the sort of index to be used as the basis for adjustment of value; the second is the acceptability of the device; and the third is the scope of application of value linking in an economy.
Let us consider the sort of index that should be used for value linking. Any index used for this purpose must be fair and equitable in the sense that it provides a good measure of the purchasing power of money and does not favor one party over another. It must, in other words, be based on a price or sample of prices which represents the general price level and which moves in line with movements at this level. Any other representation of price rises would by definition be biased and thus not the ideal one for this purpose.
The most commonly proposed indices for value linking have been the price of gold or some foreign currency, an index based on the selling price (or average of selling prices) of the borrower, a wage index, and a general price index. Clearly, in theory the ideal one for value linking is the general price index, because any other index, by being particular, suffers from the danger of deviating from the general trend. Gold or foreign currencies would serve as a proper index if the inflating country’s currency were devalued continuously in line with the fall in its purchasing power. In most countries, however, the devaluation is undertaken periodically and in discrete jumps, so that in the interim periods the rate of exchange does not provide an accurate indicator of value. Furthermore, the rate of exchange is also affected by speculative and other pressures not directly related to internal inflation.
Similarly, any index based on the borrower’s prices, though it may seem acceptable to the borrower, suffers from the same risk—particular prices may easily deviate from the general price level because of special pressures and accidental factors—and is also liable to use as a hidden subsidy to the borrower. For instance, it has been recommended in various instances that the borrower should pay less than the true value of the loan if his selling price (and thus his income) rises less than the general price level, and that he should pay only according to the rise in the general price level if his selling price rises faster. This is patently unfair where his price is freely determined by market forces, because if his price rises less than others his debt is reduced in real terms, while if it rises more his income is increased. In the first case, there is no good reason to reduce his debt, because the relative fall in his real selling price may be an economic indication that he should reduce his output; in stable conditions, no one would recommend that debts be reduced for borrowers whose prices have declined. It would be uneconomical to subsidize his borrowing in such a case. If the price to the borrower is controlled in some way, as for some agriculturalists, it seems at first sight equitable that his debt be reduced—after all, his income is less than it would otherwise be, for reasons beyond his control, and it is not kept low as a means to discourage production. The problem is that reduction of indebtedness may not be the best way to subsidize such a producer, first, because it would lead to a misallocation of credit in the agricultural sector by its very cheapness (so that it might be used for consumption), second, because it would discriminate against savers, and third, because it is a hidden subsidy which does not serve to encourage production. The ideal solution would be to require a full repayment of debt and interest while directly subsidizing the producer’s income by grants or special buying schemes.
A wage index, though not as restricted as a specific price index, is still unsatisfactory relative to a general price index, because it is susceptible to special pressures that may cause it to diverge from the latter. It does, however, possess a special appeal to small savers in urban areas, because it ensures them that their savings maintain the same value as their income. For this reason it may be especially appropriate to value linking of mortgage payments for workers whose wages are subject to government regulation.
The general price index, as we have termed it, can be one of three indices—the cost of living index, the wholesale price index, or the price index of the gross domestic product (GDP)—each of which can give different results. Of them, the last is the most general, comprising all items that have a market value in a country. The cost of living index is based more on consumer goods and services and thus has more relevance for the savers; the wholesale price index is based more on producer goods and is thus more relevant to the borrowers. In strict theory, only the GDP price index is admissible as the perfect index for value linking, but the others may also be used, because of their special relationship to one party or another, without any significant distortion in the financial contract.
The theoretically satisfactory solution is not, however, the one which is best suited to practical application. First, all general indices, and particularly cost of living indices, suffer from the fact that they are based on data collected from the big cities or weighted to reflect the value of money to certain classes in these cities; they are not, therefore, particularly relevant to the value of money in rural areas. Intermediaries dealing in agriculture will find these indices unsuitable for their operations and will perforce have to rely on less general indices, such as individual, or collections of, agricultural prices. Second, the GDP index is impractical for value linking because, unlike the cost of living and wholesale price indices, it is compiled once a year and published many months after the relevant period. This really rules it out for our purpose, because what is needed here is a readily available monthly index of prices.
The second problem, the acceptability of the index chosen, also raises various difficulties for general indices. First, the construction of an index is a difficult and complicated task and may even be liable to deliberate distortion. If for some reason the general indices existing in an inflationary country are mistrusted by the population, a specific index, such as the price of gold or the borrower’s selling price, will have to be used instead. A lot depends on the prestige and impartiality of the agency entrusted with index compilation. It would be unfortunate if no general index could be used, because a multiplicity of particular indices can clearly bring about a great deal of confusion in financial markets and create distortions and inequities of its own.
Second, even if some general price index were available and trusted by the parties concerned, full monetary correction may prove unacceptable to the borrower for the reason that inflation increases considerably the riskiness of any proposed enterprise and some inducement in the form of less-than-full correction (say, 25 per cent correction when inflation is 30 per cent) may be required. This form of correction gives some security to borrowers who fear that their particular prices may not rise as fast as the general price level for reasons unconnected with underlying market forces. In effect it is similar to lowering interest rates in stable conditions as an encouragement to investors.
Thus, general indices of cost of living or wholesale prices may often be used for value linking in urban and industrial areas, but with less than full correction in some cases where borrowers find the risk of full correction too great. In rural areas particular price indices may be more practical, but care should be taken to make these indices as equitable as possible to both parties, and to limit the number of indices used. In the agricultural sector itself those indices can be used which are the most general and representative of the value of money.
The third problem concerns the scope of application of value linking. If value linking is introduced in one sector and not in another, borrowers in the former will feel discriminated against in favor of the latter. Correcting for inflation in particular sectors may thus create dissatisfaction in those sectors and may distort the pattern of investment in the economy. The best solution, though again not the most practical, would be to introduce value linking in all sectors together. Failing this, it may be necessary to suffer some distortion and dissatisfaction for the benefits resulting from value linking by introducing it whenever feasible, possibly by requiring government-owned intermediaries to practice it in all their transactions.
REAPPRAISAL OF A ROAD PROJECT IN IRAN
Occasional Paper No. 7
By Herman G. van der Tak and Jan de Weille
In 1959 the International Bank for Reconstruction and Development made a loan to the Government of Iran to assist in financing the road program in Iran’s Second Seven-Year Development Plan. In an effort to evaluate the actual economic effects of this project—and to improve techniques of project analysis and appraisal—the Bank undertook this reappraisal of the project it financed in 1959. After reviewing the general transport setting and historical background for the project, the authors discuss construction costs and delays, the structure and growth of traffic, road user savings, development benefits that were expected to result, and the impact of the road on agriculture. Finally, the authors proffer a cost-benefit analysis of various road sections in the project, the possible effect of the results of changes in the estimate of benefits, and the opportunities for alternative investment.
Herman G. van der Tak is author of The Economic Choice between Hydroelectric and Thermal Power Development, No. 1 in this series. Jan de Weille is author of Quantification of Road User Savings, No. 2 in the series. 152 pp./$3.00 paper Distributed for the World Bank by
THE UNITED KINGDOM
Oxford University Press, Ely House, 37 Dover Street, London, W.l, England
CONTINENTAL EUROPE, AFRICA, THE NEAR AND MIDDLE EAST, PACIFIC ISLANDS, LATIN AMERICA
International Book Export Group, P.O. Box 231, Princeton, New Jersey 08540, U.S.A.
Centro Interamericano de Libros Academico (CILA), 31 Sullivan-Bis, Mexico 4, D.F.
SOUTH AFRICA, INDIA, PAKISTAN, MALAYSIA, AUSTRALIA, NEW ZEALAND, HONG HONG
International Book Export Group, P.O. Box 231, Princeton, New Jersey 08540, and Oxford University Press, Ely House, 37 Dover Street, London W. 1, England
United Publishers Services, Ltd., Shimura Building, 1, 4-chome Kojimachi Chiyoda-ku, Tokyo, Japan, and International Book Export Group, P.O. Box 231, Princeton, New Jersey 08540, U.S.A.
The Johns Hopkins Press Baltimore, Maryland 21218, U.S.A.
The problems of financial intermediaries in an inflating economy are very difficult ones, and attempts to solve them present both a general danger of applying palliatives in a situation where more attention should be devoted to finding a cure and also pitfalls and complexities which, in a short space, can only be indicated. Yet it may be helpful to sum up the outline given here.
Inflation creates serious problems for the establishment and success of financial intermediaries in developing countries by reducing the supply of monetary saving and distorting the pattern of investment, and must in some way be counteracted if these valuable institutions are to be utilized for economic development. The mere raising of interest rates is not a sufficient remedy, because it does not compensate accurately for future price rises, which are inherently unpredictable, and because it does not prevent a shortening of lending periods. The best solution is probably to adjust lending terms to actual past inflation in a manner that maintains the real value of the loan and charges a positive and appropriate rate of interest. This is the closest one can get to reproducing the conditions of price stability in an inflationary environment.
The adjustment of loans to past inflation requires choosing an index, agreeable to both parties, which is a good representation of the decline in the value of money. Any index is better than having no adjustment at all, but some indices are more liable to bias than others. In general, the broader the coverage of the index the more it is likely to be fair and equitable, the narrower its coverage, the more likely it is to favor one party or the other—the former is to be preferred, subject to the condition that the general index is properly constructed and can win acceptance from all concerned.
In actual practice the use of value linking is likely to face various problems and to be subject to many compromises. Governments may not wish to adopt it at all if they wish to give the impression of seriously battling inflation. If they do wish to adopt it, they may have to choose different indices for different sectors, to introduce it in slow and difficult stages, and to create some dissatisfaction among the borrowers. Nevertheless, for the success of financial intermediaries, such efforts may be worthwhile, for without them they would be hopelessly dependent on government subsidy and unable to ensure an economical allocation of their financial resources.