RBI needs to be commended for several ingenious and bold steps that it has taken recently to address market microstructure and improve communications with market by engaging in constant interactions. Recently, RBI expressed its intent to inject liquidity through swap transactions. While this move has been welcomed by the markets, it has also raised an important question: How far the existing liquidity management framework is an adequate reflection of liquidity?
We estimate that all scheduled commercial banks currently have only 2.55% of excess SLR (0.55%, if we net out the G-Secs for daily transaction and cash needs) that they can use as a collateral to borrow from RBI or that can be used by RBI for OMO purchases. Hence, the question ‘why did RBI have to take the swap route for injecting durable liquidity?’ arises
The issue of not having adequate collateral brings us directly to the current working of RBI’s liquidity management framework as it is exactly based on this premise. Specific banks (mostly having a strong retail franchise) can borrow from RBI only against their collateral (that is less than 2% of bank liabilities), and there are instances where banks (wholesale) cannot borrow and must specifically mobilise public deposits to meet their funding requirement. Thus, it is no wonder that incremental credit deposit ratio is running more than 100% now, with CD and repo rate spread at more than 150 basis points. Thus, the existing liquidity framework may not be a complete indicator of liquidity tightness in the system. The logical corollary of this argument is then “Why call rates are not showing up in this liquidity tightness?”
Before we explain such an anomaly, to be fair to RBI, in the February 2019 issue of the RBI bulletin (Contours of Liquidity Management: Developments During 2018-19), there is a succinct summarisation of the liquidity operations by RBI during FY19. The article articulates how RBI managed the liquidity in the situation where fears of global trade tensions intensified and faster-than-anticipated normalisation of the US monetary policy led to capital outflows that exerted depreciation-pressure on rupee. We believe the market may be well advised to take cognizance of such communications from RBI to understand the nuances of liquidity management.
We believe the lacunae in current liquidity management by RBI is that, presently, frictional liquidity injection (repo transactions to compensate for government cash balances) is substituting for durable liquidity (injections through OMO to compensate for currency leakage and liquidity impact of RBI forex intervention), and, hence, this is resulting in an imbalance between effective mix of durable and transient liquidity injection resulting in market imperfections.
Experience suggests that the provision of short term/frictional liquidity does not substitute fully for durable liquidity, though durable liquidity can substitute for short term/frictional liquidity needs (RBI monetary policy statement: April 2016). For example, when the currency leakage was combined with the liquidity outflows/forex sales which took place this year, amid the FII outflows, then RBI was able to just compensate 51% through the durable liquidity method till January 2019. On the other hand, till December 2018, the total amount of repo injections has more than fully compensated the variation in government cash balances.
Now, coming to the weighted average call rate (WACR), over the years, the share of call money has declined significantly (currently around 10% compared to 60% share of TREP market and remaining 30% of market Repo). Additionally, the movements in WACR could be liquidity-agnostic as it has hardly moved even when deficit has been as large as `2.6 trillion. The WACR is highly skewed. In the morning (the first hour of trading in the inter-bank call money market usually accounts for about 75-80% of the day’s volume), the call rate is high, and in the evening, it is low as some banks dump their excess liquidity in the call market after netting their positions in other markets. This is a regular feature and distorts the WACR. Also, most of the co-operative banks are not participants in the NDS-Call trading platform. The absence of uniform market hours across all money market segments which are not in sync with RTGS timings often have a destabilising impact on the WACR.
Clearly, issues relating to market microstructure needs to be addressed to have a holistic assessment of the WACR being used as a proxy for liquidity management tool.
So, what next? We believe RBI could also use, as risk spreads, incremental credit deposit ratio to have a comprehensive assessment of systemic liquidity and not only call rate. First, we suggest that banks may be allowed to use the full mandatory SLR for maintaining or allowing CRR in the computation of HQLA. This will release around `5 lakh crore worth of G-secs into the system that could be used for on-lending. Second, the unspent cash balance of the government is now being auctioned by RBI through repos. Such cash balances of government can become a part of permanent liquidity if transferred to banking system and can be put to productive use. It will also provide a clear picture of the money available within the system, which will not get distorted by government borrowing. Third, can we think of a futures market for the uncollateralised call rate trading that would give an indication of future market expectation on rates as in the US?
Before we end, one suggestion for markets. In recent times, RBI communication has shifted from speeches to more rigorous research articles that underlines subtle policy changes. The market could take a cue from such publications and, thereby, tactically be in sync with RBI policy shifts. This could in effect reduce volatility in financial markets.
The author is Group Chief Economic Advisor, State Bank of India (Views are personal)