BFSI The impending change in banks base rate formula The RBI is shortly expected to announce the new base rate formula, linking banks base rates to their marginal cost of funds. This is notwithstanding the banks protest that the marginal cost of funds does not reflect the nature of their balance sheets. However, we expect the RBI will provide some relief to banks on transitional issues such as a step change in the base rate due to the change form average cost of funds to marginal cost of funds. Whilst the banks moving from BPLR to the base rate regime itself was negative for banks NIMs, the new base rate formula will make banks NIMs even more sensitive to changes in policy rates. Banks with well-matched assets and liabilities are better placed to avoid volatility in margins, but banks with a high share of floating rate loans, high CASA and slow growing CASA deposits (e.g. SBI and PNB) will find it harder to protect margins in an easing rate cycle. After the hue & cry, finally the real thing The RBI is likely to issue final guidelines on the computation of banks base rates by end-november 2015. Earlier on 1 September 2015, the RBI had released draft guidelines (https://goo.gl/axsx5a) proposing that banks base rates be linked with their marginal cost of funds. In this note, we discuss what could be the probable shape of final guidelines and the impact of these guidelines on bank s NIMs. First, here is a quick recap on regulatory developments on the base rate: Exhibit 1: History of regulations on the base rate Date Regulatory Change Banks were asked to switch over to the system of Base Rate from BPLR (Benchmark July, 2010 Prime Lending Rate) to reference loan pricing. RBI clarified that banks have discretion to use either average or marginal cost of funds January, 2015 or any other combination. If card rate of certain tenor of deposits are used, that tenor should have the largest share in bank s deposit base. Draft base rate guidelines proposed that banks use marginal cost of funds as input to September 2015 base rate calculation starting from 1 April 2016. Source: RBI, Ambit Capital research The proposed base rate computation formula is different from the existing formula in two major ways. Exhibit 2: The proposed changes in the base rate formula Existing Formula Banks have discretion to use either average cost of funds or marginal cost of funds or a blend of these two methods. Banks have to use the deposits bucket, with the largest share in total deposits, as a basis for calculation of cost of funds (average or marginal). Source: RBI, Ambit Capital research Proposed formula Banks have to necessarily use marginal cost of funds to calculate the change in the base rate Entire funding base (excluding equity) to be used for calculating the cost of funds NEGATIVE Quick Insight Analysis Meeting Note News Impact What bankers are complaining about? Since the draft guidelines were made public, the banks have articulated three main complaints about the proposed guidelines: Where marginal cost of funding is materially lower than average cost of funds currently being used by a bank, there could be significant one time impact simply due to a transition to using marginal cost of funds. The marginal cost of funds does not reflect the cost of funds that banks bear on their balance sheets, as deposits re-price with lag. However, linking the base rate with marginal cost of funds will lead to immediate re-pricing of floating rate loans, following any change in card rates of deposits. Thus, banks net interest margins could be significantly impacted negatively during the easing rate cycle. Analysts Ravi Singh Tel: +91 22 3043 3181 ravisingh@ambitcapital.com Pankaj Agarwal, CFA Tel: +91 22 3043 3206 pankajagarwal@ambitcapital.com Aadesh Mehta, CFA Tel: +91 22 3043 3239 aadeshmehta@ambitcapital.com
The use of the entire funding base (including CASA) instead of a particular tenor bucket, for calculation of cost of funds will lead to a situation where cost of funds could be much lower than 364-days Treasury rates (used as a benchmark for return on SLR investment). As a result, it may lead to a situation where there is no negative carry for SLR & CRR. What the RBI is likely to do? We believe that the RBI will go ahead with changing the base rate calculation and link it to the banks marginal cost of funds and whilst, in the final guidelines, the RBI is likely to address banks concerns on complaints number 1 and 3, it is unlikely to agree with complaint number 2. We believe so because the potential step changes in base rates as highlighted in complaints 1 & 3 are mathematical in nature and the RBI could provide either a transition mechanism to smoothen the impact over a longer time period or allow banks to flex, for once, other components in the base rate formula (e.g. increase the spread for a desired RoE) to offset the impact of a one-time change in the measure of cost of funds. On the more fundamental objection (complaint 2) of banks - that the nature of banks balance sheets (lagged re-pricing of deposits and immediate re-pricing of floating loans) impedes the use of marginal cost of funds - the RBI is unlikely to give the banks any leeway. For example, in the Q&A following the 29 September monetary policy, the RBI governor noted: banks have typically become used to a reasonably slow moving base rate which gives them time to adjust the deposit costs before they adjust lending the bottom line is the market pressures and competitive pressures from a fast moving bank like HDFC [Bank] would force other banks to move towards a more marginal cost based rate what western banks do, clearly, they use asset/liability management to make sure that they do not have a huge mismatch between the interest rate sensitivity of their assets and the interest rate sensitivity of their liabilities. But also if there is still a residual sensitivity they use the derivative markets to balance that sensitivity so that they are hedged. I think our banks will have to move towards that. You cannot continue protesting that your costs are moving but the incremental loans require faster movement Thus, with transmission of monetary policy being main concern of the RBI, the RBI is likely to go ahead with necessary linking of banks base rate with marginal cost of funds, which will ensure that banks base rates react immediately to policy rates. What would be the impact on banks? Even before we go into the implications of using a base rate based on marginal cost of funds, we wish to highlight that an easing rate cycle for first time under a base rate mechanism (in contrast to the BPLR mechanism) in itself was fairly negative for banks NIMs as explained below. Banks will face base rate regime in an easing cycle for the first time: The BPLR mechanism allowed banks to not cut BPLR in a declining rate environment and yet attract new borrowers by offering lower rates at discount to BPLR. Thus, the banks did not pass on their lower cost of funds by reducing their BPLRs in a declining rate environment as compared to quickly raising BPLRs in a rising rate environment. For example, SBI s PLR in at 13.0% was 150bps higher than in FY01 vs a 200-250bps decline in SBI s funding cost, the repo rate and the 10-year Government bond yield during this period. The RBI may provide some relief to banks to absorb transitional impact of change in base rate formula. Banks have typically become used to a reasonably slow moving base rate With transmission of monetary policy being main concern of the RBI, it is likely to go ahead with marginal cost of funds based base rate formula. Base rate mechanism, in itself, had made things trickier for banks.
Exhibit 3: Change in SBI s BPLR, cost of funds and market rates over FY01-11 End-FY01 End- Change SBI PLR 11.50% 13.00% +1.50% SBI Cost of funds 7.62% 4.98% -2.64% Repo rate 8.75% 6.75% -2.00% 10 year G-sec yield 10.32% 7.99% -2.33% Once the RBI introduced the base rate, the impact on NIMs has been visible, for example, in the way SBI s base rate and BPLR have moved after the introduction of base rates. During -14, the increase in the base rate has been in-line with an increase in the cost of funds. Exhibit 4: PLR and base rates have moved in line with borrowing costs after the base rate mechanism was put in place FY01-05 F05-09 -11-14 Change in SBI s average PLR/base rate -161bps +251bps -49bps +18bps Change in SBI s average cost of funds -252bps +88bps -99bps +22bps However, now for the first time, the impact of the base rate mechanism will be visible in a downward cycle. The recurring impact of new base rate formula is high sensitivity of NIMs to policy rates: Whilst draft guidelines indicate, a significant one time shift in base rates due to the change in methodology, as discussed above, we expect the RBI to provide some relaxations to banks on transitional issues arising from a change in the base rate formula. However, despite these relaxations, even on an ongoing basis banks; NIMs should become highly sensitive to policy rates and should moderate in a declining rate cycle. The sensitivity of banks NIMs to policy rates would be the result of residual sensitivity between assets and liability side sensitivities. For example for SBI, on the assets side, its yield on loans is highly sensitive to a change in PLR/base rate (as a reflection of repo rate). This is because almost 90-95% of SBI s loans are base rate linked. The correlation between annual change in PLR and SBI s yield on loans is 90%. Exhibit 5: SBI s yield on loans is highly correlated to the bank s PLR/base rate On an ongoing basis, banks NIMs would become more sensitive to policy rates and should moderate in a declining rate cycle. Change in PLR Change in yield on advances 2.5% 2.0% 1.5% 1.0% 0.5% 0.0% -0.5% -1.0% -1.5% FY02 FY03 FY04 FY05 FY06 FY07 However, on the liabilities side, SBI s cost of deposits are relatively less sensitive to changes in card rates (marginal) on term deposits as old deposits mature with a lag. The correlation between annual change in 6 months term deposit rate and cost of deposits is just 51%. Notably, the correlation improves to 69%, if card rates with 1- year lag are used.
Exhibit 6: Low correlation between SBI s card rate of deposits and cost of deposits in the same period Exhibit 7: SBI s cost of deposits are more closely correlated with card rates on 1-year lag basis. Change in TD rate - 6 months Change in cost of deposits Chg. In 6 mths. TD rate, 1 yr lag Change in cost of deposits 2.0% 1.5% 1.0% 0.5% 0.0% -0.5% -1.0% -1.5% -2.0% 2.0% 1.5% 1.0% 0.5% 0.0% -0.5% -1.0% -1.5% -2.0% FY02 FY03 FY04 FY05 FY06 FY07 FY03 FY04 FY05 FY06 FY07 Thus, due to SBI s assets yield being highly sensitive to base rate/plr, it had historically made more sense for SBI to link its base rate calculation with average cost of deposits rather than the card rates on deposits (marginal), to minimize the volatility in spreads. However, the regulatory pressure and market forces mean that the offset is not complete and, thus, despite base rate/plr being linked to average cost of deposits, SBI s spreads/nim show a residual sensitivity with policy rate, i.e. rise in tightening cycle and fall in easing cycle. Exhibit 8: SBI s NIM (domestic) have followed the direction of change in policy rates 8.1% 4.3% 7.6% 4.1% 7.1% 3.9% 6.6% 3.7% 6.1% 3.5% 5.6% 3.3% 5.1% 3.1% 4.6% 2.9% Avg Repo rate SBI NIM (RHS) Thus, if the RBI makes base rates linked to the marginal cost of funds, SBI s base rate will swing even more with card rates (marginal) on deposits. Whilst, it is difficult to model this sensitivity, it is a cyclical phenomenon and over the cycle, the impact of rate cycle should be neutral (all else being unchanged). However, early in an easing cycle, NIM seems highly likely to compress. Assuming a 50bps cut in card rates on term deposits, SBI s marginal cost of funds will change by ~30bp (assuming 40% CASA/funds). Thus the base rate will change by 30 bps. However, assuming 2 years average maturity, cost of deposits will change by just ~15bps. Thus, in the first year alone, after a 50bps policy rate cut transmission, spread mays decline by ~15bps. This 15bps impact is due to the sensitivity of NIMs to the rate cycle alone. Other factors which would could mitigate or accentuate this impact would be a change in liability mix, change in the asset mix and intensity of competition in the key lending segment of the bank.
In general, four factors that determine the sensitivity of banks NIMs to policy rate cycle. They are: (i) share of floating rate loan in total loans; (ii) share of CASA deposits; (iii) changes in mix between floating vs fixed rate loans; and (iv) growth in CASA compared to balance sheet growth. In an easing rate cycle, the bank which stands to benefit is the one with: (i) higher share of fixed rate loans; (ii) lower CASA ratio; (iii) fixed rate loans growing faster than floating rate loans; and (iv) faster growth in CASA deposits. Thus, whilst different banks are placed differently on these parameters, IndusInd Bank ticks the boxes on all four factors, whilst PSU banks, such as State Bank of India and Punjab National Bank are adversely placed on all the four factors.
IndusInd Bank Ltd (IIB IN, BUY) 1,200 1,000 800 600 400 200 0 Nov-12 Jan-13 Mar-13 May-13 Jul-13 Sep-13 Nov-13 Jan-14 Mar-14 May-14 Jul-14 Sep-14 Nov-14 Jan-15 Mar-15 May-15 Jul-15 Sep-15 Nov-15 INDUSIND BANK LTD Source: Bloomberg, Ambit Capital research Punjab National Bank (PNB IN, SELL) 250 200 150 100 50 0 Oct-12 Dec-12 Feb-13 Apr-13 Jun-13 Aug-13 Oct-13 Dec-13 Feb-14 Apr-14 Jun-14 Aug-14 Oct-14 Dec-14 Feb-15 Apr-15 Jun-15 Aug-15 PUNJAB NATIONAL BANK Source: Bloomberg, Ambit Capital research State Bank of India (SBIN IN, SELL) 350 300 250 200 150 100 50 Nov-12 Jan-13 Mar-13 May-13 Jul-13 Sep-13 Nov-13 Jan-14 Mar-14 May-14 Jul-14 Sep-14 Nov-14 Jan-15 Mar-15 May-15 Jul-15 Sep-15 Nov-15 STATE BANK OF INDIA Source: Bloomberg, Ambit Capital research