The Reserve Bank of India (RBI) is switching from a Marginal Cost of Funds based Lending Rate (MCLR) system to an external benchmark based lending rate system from April 1, 2019. This shift in policy was announced via RBI’s ‘Fifth bi-monthly monetary policy statement for 2018-19’ and was based mainly on ‘The Report of the Internal Study Group to Review the Working of the MCLR System’ which in turn was released for public feedback on October 4, 2017.

 A quick remark on the excellent process of policy-making at the RBI seems to be due here. For instance, a major policy shift (in this case a shift out of the MCLR system), involved detailed study of the matter by an expert panel and dissemination of findings (a 121 page report) among the public for feedback with adequate time between circulation and final policy decision.

However, an internal benchmark based lending rate system especially one such as the ‘base rate system’ seems superior to not only other internal benchmark based lending rate systems such as the MCLR system but also to external benchmark based lending rate systems on logical (a priori), empirical (a posteriori) and practical (psychological & behavioral) grounds.  This article details argument and analysis to justify this fact.

I. What are internal and external benchmark based lending rate systems?

Banks are in the business of financial intermediation, i.e. mobilizing funds from the surplus sector and channeling it to economic agents who need these funds. Interest rates are prices which banks pay to depositors and creditors for these funds as well as prices which they receive from borrowers for use of these funds.

Benchmark based lending rate systems seek to make pricing of bank loans organized and transparent. Benchmarking seems especially important for free market economies in which interest rates are market determined because benchmarking supplies key decisional information (about price setting) to economic agents, thereby leading to (near) perfect information – the lynchpin condition required for efficient functioning of a market.

However, internal and external benchmarking systems differ in an important way. While the former is based on, among other things, internal cost of funds, the latter is based on key external interest rates such as the central bank’s policy rate for instance. Hence, external benchmarks unlike internal benchmarks are outside control of banks.

Internal benchmarks can be of various types. For instance, the RBI had introduced Prime Lending Rate (PLR) as the internal benchmark for banks in year 1994. ‘The Report of the Internal Study Group to Review the Working of the MCLR Based Lending Rate System’ describes PLR as “the interest rate charged for the most creditworthy borrowers taking into account factors such as cost of funds and transaction costs.” PLR in India was expected to act as a floor lending rate for loans above INR 2 lakhs. The RBI eventually moved out of the PLR system in year 2003 because of wide variation in PLR, wide variation in spreads and because PLR of banks turned out to be quite inflexible with respect to the general direction of interest rates in the economy.

Benchmark Prime Lending Rate (BPLR) is another type of internal benchmark. The RBI introduced BPLR into the banking industry in April 2003 after dissatisfaction with its PLR system. The RBI internal study group report states that under this system, banks were required to announce their BPLR considering cost of funds, operational costs, minimum margin (to cover regulatory requirements) and profit margin. This system too was dumped in July 2010 due to problems such as excessive sub-BPLR lending (apparently as high as 77% in September 2008) such that sectors outside the purview of the BPLR system were effectively cross-subsidized at the expense of sectors covered by the BPLR system.

The RBI then introduced the base rate system, another internal benchmark based lending rate system, in July 2010. Base rate comprised of the following: i) cost of borrowed funds, ii) negative carry on cash reserve ratio (CRR), iii) unallocatable overhead cost and iv) average return on net worth. Banks were required to announce their base rate which was to be the minimum rate for all loans covered by this system. Actual lending rate also incorporated a borrower-specific premium. The RBI dropped this system in year 2016 because “it led to opacity in determination of lending rates by banks and clouded an accurate assessment of the speed and strength of (policy) transmission.”

Nepal Rastra Bank (NRB), the central bank of Nepal, also introduced the base rate system for commercial banks through its Monetary Policy for FY 2012/13. At that time, base rate consisted of following components: i) cost of funds, ii) cost of CRR, iii) cost of Statutory Liquidity Ratio (SLR), iv) cost of operations and, v) return on assets (ROA). At present, the base rate system has been expanded to cover all banks and financial institutions. Also, the NRB has recently removed (through its circular dated December 26, 2018) the ROA component from calculation of base rate.

MCLR system is yet another variation in the internal benchmark based lending rate system to which the RBI moved on from April 2016. The MCLR system adopted by the RBI states following as the components of a bank’s MCLR: i) marginal cost of funds (92% of marginal cost of deposits & other borrowings +8% of return on net worth calculated through any pricing model such as the Capital Asset Pricing Model (CAPM), ii) negative carry on CRR, iii) operating cost and, iv) tenor premium or discount. Actual lending rate in this system is the MCLR plus a spread which captures business strategy and credit risk premium as maturity risk arising from longer duration loans are already factored into MCLR via tenor premium/discount.

The RBI has recently (on December 2018) decided to move on from the MCLR system to an external benchmark based system starting from April 1, 2019. This decision was largely based on findings of the ‘Report of the Internal Study Group to Review the Working of the MCLR Based Lending Rate System’ which states poor monetary transmission under the MCLR system as a key reason behind this move. For instance, the report finds that monetary policy transmission has been slow and incomplete under both the base rate and the MCLR system. Similarly, it reports that transmission has been significant on fresh loans but muted for outstanding loans, it has been uneven across borrower categories and has been asymmetric over monetary policy cycles such that transmission has been higher during tightening phase and lower during easing phase.

II. Superior monetary transmission – the chief argument for external benchmark systems

The European Central Bank (ECB) defines transmission mechanism of monetary policy as the process through which monetary policy decisions affect the economy in general and price level in particular. Standard literature on monetary transmission describes various channels such as interest rate channel, asset price channel, credit channel, etc through which monetary policy decisions affect the economy. For instance, Frederic S. Mishkin’s (of the Federal Reserve System) working paper entitled ‘The channels of monetary transmission: Lessons for monetary policy’ published in the NBER Working Paper Series in February 1996 provides an excellent basic insight into these transmission channels.

The key starting point, common across all channels of monetary transmission, is the impact of monetary policy decisions on short-term nominal interest rates. For instance, an expansionary monetary policy (a cut in policy/repo rate for instance) will first lead to a decline in nominal interbank rates which will then spill over to other nominal money market rates. Mishkin argues in his paper that since prices are sticky in the short-run, this means short-run real rates will decline as well. Furthermore, the expectations hypothesis of the term structure states that long-run interest rate is an average of expected future short-term rates. Therefore, real long-term interest rates too will decline as a result of monetary expansion.

From this point forward, an explanation of how the real economy is affected varies across different channels. For instance, the interest rate channel relies on demand side effects (cost of capital) while the credit channel relies on supply side effects (funds available for lending at banks) for monetary transmission. Similarly, the asset price channel relies on other mechanisms such as ‘wealth effect’ (which determines consumption spending) and ‘Tobin’s q’ (which determines business investment spending) for monetary transmission.

Superior monetary policy transmission seems to be the chief argument for external benchmark based lending rate systems. This certainly seems to be the major premise based on which the RBI internal study group report (discussed above) recommends a shift to an external benchmark.

Relevant to this point, Dr. Viral V Acharya, Deputy Governor of the RBI, in his excellent monetary transmission paper presented in Aveek Guha Memorial Lecture (November 2016) presents an insightful analysis of the poor pass through from policy rate changes to lending rates under the base rate and MCLR systems in India. For instance, under the base rate system, the pass through to outstanding loans from the repo rate was around 60% during the tightening phase while it was less than 40% during the easing phase. During periods of monetary easing, banks were found to ‘doctor’ their base rates as well as their MCLR to prevent these rates from falling in line with the cost of funds. Banks were also found to offset decline in their MCLR by adjusting (upward) their spreads in an arbitrary manner.

However, one cannot help but notice some interesting observations in this analysis. First, the base rate system in India seems to have been designed in such a manner in India that banks could choose between average, marginal and blended costs in calculating their cost of funds. Indeed it seems that this flexibility in choice is what made the base rate system in India ‘opaque’ (as the expert internal study group report terms it), than the base rate system per se. As base rates present the logic behind pricing of bank loans by breaking down interest rates into various component costs, the base rate system seems rather the antithesis of an opaque system.

Next, external benchmarks are said to be better (more transparent) than internal benchmarks because under the former system, the benchmark interest rate is the same for all banks so that economic agents can quickly compare various loan offers simply by comparing spreads of various banks over this common benchmark. However, economic agents can always do the same under an internal benchmark system such as the base rate system, by comparing different base rates offered by various banks.

Next criticism – that of poor monetary transmission – of the base rate system in particular and the internal benchmark system in general, seems to require a more nuanced analysis. Most policymakers and economists (barring Monetarists and New Classical economists) would agree that a system that allows better monetary transmission is superior and preferable to a system that allows constrained transmission. The New Keynesian position that policymakers (and especially central bankers) have informational and capability related advantages which place them in a better position to intervene in the economy also sounds quite valid.

However, the idea that an external benchmark based lending rate system is more appropriate because internal benchmarks such as base rates and MCLR either do not reflect marginal costs or do not adequately reflect changes in marginal costs, seems not correct. The reason is simple – external benchmarks such as central bank policy rates, for instance, adequately represent neither average nor marginal costs for banks. As an example, RBI Deputy Governor Dr. Acharya’s paper referred to above states that 90% of total liabilities of Indian banks are in the form of deposits, bulk of which are in the form of term deposits. Banks could of course move interest rates on current and savings account much quickly taking cue from monetary policy stance (and apparently this has not happened as efficiently in India in the last decade as per Dr. Acharya’s paper) so as to improve policy pass through and transmission. However, even then, it seems rather natural that the pass through to lending rates is gradual and occurs with a lag because it takes time for average costs of banks (or full costs) to fully reflect this marginal change in cost of funds. This last point brings to front an important theoretical debate in economics – the full cost principle vs. marginal cost principle argument. Indeed, it is in the answer to this important historical economic debate that one can find the most convincing argument in favor of the base rate system.

III. The theory of value in economics – a brief review of economic history

A walk down the history lane of the field of economics reveals the fact that ‘the marginalist method’ made a rather late entry into the economics discipline via the economic school now known as Neoclassicism. Before that, Classical economics, the then orthodoxy, argued that because the supply curve of firms was horizontal, demand did not matter in determining price of a commodity. The great Classical economists such as Adam Smith, David Ricardo and J.S. Mill largely subscribed to the ‘cost of production theory of value’ which argues that exchange value of a commodity is determined by cost of factors of production employed in the production process.

 Karl Marx, the great 19^th century economist (who also coined the term ‘Classical economics’) goes a step further to argue in his most excellent book ‘Capital-Volume I’ that value of a commodity depends only on labor. A more nuanced analysis of Marxian economics reveals the fact that Marx factored in the cost of other factors of production (capital and land) in terms of labor hours expended in their creation/development at earlier times so that Marx’s theory of value too is essentially a cost of production theory of value with the additional (and generally invalid) argument that only labor is the rightful owner of the proceeds from production.

The marginalist method was introduced into the economics discipline by the great Neoclassical economists such as W.S. Jevons, Alfred Marshall, Carl Menger and Leon Walras. Among other things, neoclassical economics argues that the profit maximizing level of output for a firm is the one at which the firm equates marginal cost with marginal revenue. Mathematically, there is nothing wrong with this argument or the marginalist method. However, empirical evidence has consistently gone against marginal cost pricing and in favor of cost of production theory of value (or full cost pricing).

IV. Full cost pricing vs. marginal cost pricing – empirical evidences

The first major empirical work in this regard investigating pricing strategies and other decisional processes of firms in the real world seems to have been undertaken by R.L Hall and C.J. Hitch, the findings from which are known today as ‘The Hall and Hitch Report.’ This report finds, among other things, that firms set their prices based on the full cost principle (average cost plus a ‘normal’ profit margin) rather than using the mathematically rational marginalist method.

Subsequent studies conducted after this seem to have consistently shown that firms employ full cost pricing strategy and not marginal cost pricing strategy in real life. The reasons for this seemingly irrational, value destroying behavior of firms are actually quite straightforward: i) Firms do not adequately know their demand curve to employ marginal cost pricing strategy, ii) Full cost pricing is much easier to comprehend and employ.

Jacob Gramlich of the Federal Reserve System and Korok Ray of Mays School of Business state the former as an explanation of why firms are consistently seen to choose full cost pricing in their 2015 research paper entitled ‘Reconciling Full-cost and Marginal Cost Pricing.’ They argue in their paper that a lower informational burden pushes firms towards the more practical alternative of full cost pricing. The second reason is probably best explained from the point of view of behavioral economics.

V. Thalerian behavioral economics and economic decision making

Richard Thaler won the 2017 Nobel Prize in economics for his work on behavioral economics.  ‘Nudge’, an interesting book by Thaler and Cass Sunstein presents scientific insights into the process of economic decision making by people in the real world. In this context, the book introduces an interesting dichotomy – Humans vs. Econs – where humans are everyday people while Econs are economists. The book argues that because most economic models are based on unrealistic assumptions such as perfect information and perfect information processing capability, these models are relevant only for a world of Econs. The world however, is filled by everyday people or Humans who, bounded by limited access to relevant information, limited information processing capability and rush of everyday life, mostly succumb to decisional shortcuts or heuristics.

For instance, most people employ what Thaler terms ‘mental accounting’ so as to earmark low interest earning money in one’s savings account for a future need rather than using that money now to pay off one’s much higher interest costing credit card bill, thereby not utilizing a clear arbitrage opportunity. People tend to go for this seemingly less optimal choice because it acts a check on the possibility of overspending, protects their savings and requires a lot less time and financial management skills on their part to employ. From the point of view of behavioral economics, it is not difficult to comprehend why firms (managed by Humans and not Econs) prefer the much simpler ‘average cost plus a profit margin’ (full cost pricing) principle to the marginal cost alternative.

VI. A more practical alternative for smoother monetary transmission

Shifting to the external benchmark alternative seems a very roundabout and unnatural way to ensuring smoother monetary transmission. Because external benchmark interest rates (e.g. policy repo rate or bank rate) do not represent average costs for banks, it seems quite unnatural for them to price their loans based on these rates.

Banks are therefore likely to get around this requirement by arbitrarily increasing their spreads as Indian banks seem to have done to get around the MCLR and base rate system as stated in the report of the expert RBI study group. This suggests that partial solution to the problem of ineffective monetary transmission under the base rate system lies in effective supervision of banks – monitoring their calculation of base rates and setting of spreads for consistency, accuracy and rationality.

Effective bank supervision should also ensure that banks do not selectively pass through change in policy rates to lending rates – another finding of the RBI internal study group report. Banks would naturally be more inclined to quickly pass through policy rate increases while being reluctant to pass through rate cuts. An external benchmark system might not solve this problem because the problem seems to lie in the domain of bank supervision.

One practical (and more straightforward) way to ensure effective transmission under the base rate system might be to require the banking sector to immediately revise its interest rates on deposits in line with changes in central bank policy rates. RBI Deputy Governor Dr. Acharya makes an excellent point in this regard in his monetary transmission paper. For instance, Indian banks seem to have kept their savings deposit rate unchanged at 4% between October 2011 and July 2017 despite the fact that policy rates had moved significantly from 8.5% in October 2011 to 6% in August 2017. Requiring all banks to revise their deposit rates in line with central bank policy rate changes might ensure much better monetary transmission. The transmission will of course only occur with a lag as it will take some time for average costs of banks to decline. However, this process would be more natural and therefore more appropriate.

VII. Conclusion

Classical economics as well as modern behavioral economics suggest that firms opt for full (or average) cost pricing rather than marginal cost pricing. Empirical evidence consistently and strongly backs up this proposition. Because base rates do a good job of representing average costs of banks, they seem to be the most appropriate choice to benchmark lending rates. Monetary transmission under a base rate rooted lending rate system can be improved by taking measures to ensure banks pass through change in central bank policy rates to the interest rate on their deposits in a timely manner. This might involve employing relatively soft measures such as conducting open market operations so as to ensure interbank rates remain within desired limits or using more direct measures such as establishing soft rules for the banking industry to revise its deposit rates in line with changes in monetary policy stance. Effective bank supervision also seems important in this regard to ensure banks calculate base rates in a consistent manner and do not arbitrarily tweak their spreads so as to affect pass through to lending rates following change in monetary policy stance. Finally, policymakers might have to accept the fact that monetary transmission under this arrangement will take place with a lag because average costs of banks will take some time to be moved. On the positive side though, banks will be able to revise interest rates on outstanding loans (and not only on new loans) more easily under this system and the process will be much more natural rather than enforced.

(Originally published in the 2024 April, i.e. 28th edition of Nepal Rastra Bank Employee Union’s publication entitled ‘Arunodaya’)