In search of a rupee liquidity framework – Part II

(This article appeared on CNBCTV18 online on April 2, 2019, per link below:

https://www.cnbctv18.com/economy/why-india-badly-needs-a-rupee-liquidity-framework-part-ii-2809951.htm )

In search of a rupee liquidity framework – Part II

In Part I yesterday, we tried to bridge different viewpoints on rupee liquidity. 

Let’s now consider the tools to manage banking liquidity, from three angles – the tenor of impact, the quantum of impact, and the side effects.

Cash Reserve Ratio (CRR)

A 1% cut in CRR would today infuse INR 1.25 tn as permanent banking liquidity. 

As a side-effect, CRR changes have a cascading, enduring impact on future liquidity. A CRR cut reduces the amount to be set aside from future deposits, allowing banks to lend more and generate more liquidity through a higher “money multiplier”.

Given the already low level of CRR (4%), and the enduring impact of CRR changes, RBI has used this tool sparingly. 

Repos, reverse repo & standing deposit facility

RBI can inject (withdraw) banking liquidity against the collateral of bonds through repos (reverse repos).

The liquidity impact is to the tenor of the repo/ reverse repo. 

The amount of liquidity injection by RBI through repos is limited by the extent of surplus bonds with banks.

RBI should continue reducing SLR and allowing more SLR bonds to be reckoned for LCR to increase surplus bonds with banks. Eventually, RBI should consider accepting good quality corporate bonds for repo and help develop secondary trading in debt capital markets. 

On the flip side, the extent of possible liquidity withdrawal by RBI through reverse repos is limited by the amount of bonds held by RBI. 

RBI has also sought a Standing Deposit Facility (SDF) so that it can withdraw bank liquidity even without providing the collateral of bonds. 

Arguably, repos have the least side effects into currency or bond markets. 

Bond Open Market Operations (OMOs)

RBI’s outright buying (selling) of bonds injects (sucks out) liquidity to the tenor of the bond. 

As a side-effect, OMOs directly impact government bond prices.

During 2017, RBI’s sale of INR 900 bn of bonds to suck out liquidity post demonetization contributed to the eventual sharp fall in bond prices. 

Conversely, in FY19, RBI purchased nearly INR 3 tn of bonds. This lifted bond prices, and RBI effectively financed 70% of the net issuance by the government. 

Foreign currency spot intervention

RBI intervenes in currency spot markets as a heat sink to counter large net inflows/ outflows of foreign currency.

Besides impacting spot market prices, RBI’s purchase (sale) of foreign currency in the spot market injects (withdraws) rupee liquidity. 

The tenor of the liquidity impact depends on how durable the underlying currency flow is. 

Foreign currency forward and swap intervention

RBI can also choose to intervene in the currency forward market to avoid any immediate impact on rupee liquidity. 

As a variation, it can undertake FX swaps as it did last week. It can inject (withdraw) rupee liquidity by buying (selling) foreign currency in the spot market, and simultaneously selling (buying) foreign currency in the forward market.

As a side effect, RBI’s forward sale (purchase) of foreign currency would decrease (increase) the forward USDINR rate, which should attract forward buyers (sellers) of foreign currency.

Currency in circulation

While Currency in Circulation (CIC) isn’t an active tool used by RBI for liquidity management, withdrawal (deposit) of currency notes decreases (increases) banking liquidity. 

Demonetization resulted in a sharp, transient increase in banking liquidity, while seasonal increase in CIC prior to elections and festivals reduces liquidity. 

RBI has to consider how much of CIC changes are permanent and how much are transient.

Government balances

This is more a liquidity irritant than a tool. Per law, government needs to bank with the RBI. All payments to the government therefore move out of the banking system into RBI, until the government spends the money back. 

While the RBI is currently undertaking repos to manage the resultant liquidity aberrations, eventually, the government should be allowed to bank with commercial banks.

Building a liquidity framework

Policy makers may need to consider three issues while managing liquidity. 

First, what is the desired liquidity status – surplus, neutral or deficit? This should depend on the monetary policy stance.

Second, if liquidity in the system needs alteration, what is the time-frame of adjustment needed?  

Third, the use of which instrument involves manageable side-effects, particularly into currency and bond markets?

The desired liquidity status

In March 2011, an RBI working group determined that the banking system should always operate in a permanent liquidity deficit mode. 

This endured till April 2016, when then RBI governor Dr. Rajan listened to some shrill market voices and finally moved liquidity to a neutral mode, to improve transmission of policy rate cuts.

Liquidity status is a powerful adjunct to transmit a monetary policy stance, and in my view, all options – surplus, neutral and deficit – should be kept open. 

A policymaker should never say never.  

When the banking system is maintained in surplus liquidity, as it was post demonetization, banks are forced to park a part of their deposits with the RBI at low yielding reverse-repo. This goads individual banks to focus on buying bonds and lending money, rather than raising deposits. This transmits lower interest rates into the system. 

Conversely, when the banking system is kept consistently short of liquidity, individual banks focus on raising deposits and selling bonds rather than giving out loans. Much like the mule chasing the carrot at the end of the stick, as long as liquidity remains in deficit, banks’ desire for deposits is never satiated. This hunger transmits higher rates into the system.

The choice of keeping liquidity in constant surplus has to be made carefully. There have to be macroprudential controls to ensure that surplus liquidity is not used as a palliative to prop up unviable assets and postpone tough executive reforms, only to eventually risk greater financial instability. The possible impact on areas such as household savings have also to be considered carefully.

Determining the time-frame of the liquidity solution

The tenor of impact of the liquidity tool used has to be appropriate to the context. 

As an example, post demonetization, banks were saddled with surplus liquidity for about a year. By conducting INR 900 bn of OMO bond sales in 2017, RBI withdrew some of the liquidity on a durable basis, besides pumping in bond duration into the system. Instead, 1-year term reverse repos would have perhaps been far more appropriate during the period.  

On the flip side, during FY19, RBI sold around US$26 bn in spot markets to protect the rupee. This led to a sharp withdrawal of rupee liquidity. To counter this and the impact of higher CIC, RBI conducted OMO purchases of INR 3 tn. One could argue that some of the reduction in foreign currency reserves should have been treated as transient, perhaps over a period of 2 years. A chunk of the OMO purchases during FY19 could perhaps have been replaced by long term repos.

Understanding the side-effects of liquidity tools

As any trader will tell you, monetary policy, liquidity, currency and bond markets are interconnected. In choosing liquidity tools, policymakers have to ensure that the side-effects into other areas are acceptable. Here are a few examples worth reviewing:

·     The use of FX forwards FY15-18

During FY15-18 (April 2014 to March 2018), RBI purchased USD 52 bn in currency forward markets, and avoided the rupee liquidity injection that accompanies spot purchases. In turn, rupee liquidity was in deficit during much of the period, keeping overall rates relatively high. 

As a side-effect, these forward purchases kept USDINR forwards high. This attracted even more foreign currency inflows in the form of rupee carry trades - including net exporter selling, unhedged ECBs, FPI in debt and outright speculative INR long positions. These caused overvaluation of the rupee, hurt exports, and left the system vulnerable for an eventual sharp reversal. 

It is ironical that during the period, there was much angst around lack of transmission of rate cuts.

It would have perhaps been advisable to restrict all currency intervention during FY15-18 to the spot market, allow the system to go into rupee liquidity surplus, allow USDINR forward premia and short-term yields to come down, arrest reversible USD inflows, and reduce rupee vulnerability. The surplus liquidity would have also tied in with an accommodative monetary stance, and transmitted rate cuts better into the system. 

·     Use of OMO bond purchases during FY19

As mentioned earlier, RBI purchased INR 3 tn of bonds during FY19 to infuse rupee liquidity. This amounted to over 70% of the net GOI borrowing program for FY19. 

Such unprecedented large-scale OMO purchases supported bond prices and the GOI issuance program, at a time of serious fiscal ambiguity. 

At this scale, this begins to look uncomfortably like printing money to finance a fiscal deficit.

Regulators seem to have trouble acknowledging this impact. There appears to be a belief that if an effect is not intended, it is not caused. 

Any liquidity framework has to debate the desirability of such a large side-effect. This has to be an eyes-open decision, rather than the blinkered following of a rule. 

Personally, I would argue that a significant chunk of the OMOs could have been replaced by long-term repos out to 1-year or more. These would have provided liquidity of sufficient tenor while avoiding distortion of the GOI bond curve.

·     Use of FX swaps last week

Last week, RBI injected 3-year rupee liquidity by buying USD in the spot market, and simultaneously selling USD 3-years forward.  This 3-year rupee liquidity infusion should help bring down the systemwide cost of funds up to that tenor.

As described earlier, the side-effect of this is the bringing down of the forward USDINR rate, making it cheaper for importers and ECB borrowers to hedge their exposures. 

With the recent inflows from FPI and FDI, INR has strengthened in 2019. However, the nature of the external balance is still brittle and in real effective exchange rate terms, rupee is overvalued. 

Under these circumstances, making it attractive for importers and ECB borrowers to hedge their exposures is a good side-effect to contend with – for now. 

However, this tool can be tricky if the fortunes of USDINR were to change. If there were sudden panicky USD outflows requiring RBI to sell USD, these RBI buy/ sell swaps would have cheapened USDINR FX forwards, and therefore made it easier for importers/ FPI and ECB borrowers to rush out and buy even more foreign currency – quite the reverse of the issue we saw during FY15-18. 

Conclusion

For liquidity management, it is important to know what the desired liquidity state is, be aware of the tools at one's disposal, be aware of the tenor of impact, and the inevitable side-effects of each tool. 

Liquidity management is intertwined with monetary policy transmission, currency markets, bond markets and macroprudential stability. No-one ever said this was simple. We simplify this for simplicity’s sake at our own risk.