The 2013 RBI FCNR(B) Swap Window – Review & Takeaways

In the wake of recent volatility in our currency markets, commentators have – somewhat prematurely - started to recall the extraordinary steps taken by RBI in 2013. 

To recap, post the Bernanke taper tantrum of mid-2013, besides other steps, the RBI announced two special swap windows in September 2013.

First, it offered to swap US$ raised by banks from foreign currency non-resident (FCNR) deposits of maturity 3-year and above into INR, at a concessional rate of 3.5% p.a. This was about 3.0% cheaper than the average market during the period.  Second, it allowed banks to borrow additional FCY funds from overseas and swap them into INR at a concessional rate of 1% below market. 

Collectively, the two swap windows brought in $34B at a crucial time for India. The first swap against FCNR(B) deposits, in particular, exceeded all expectations and raised $26B of 3-year and above money – much more than the previous schemes of India Millennium Deposits (2000) and Resurgent India Bond (1998).

To reiterate, I believe it is unnecessary for the RBI to execute any such unconventional scheme as yet. While we could see continued US$ outflows, RBI has plenty of FX reserves and our currency is still overvalued in REER terms. 

Having said that, it is always prudent to review and brush up on contingency plans - after all, what if oil prices moved up to $100/ barrel?

Specifically, two aspects of the 2013 swap window against FCNR(B) deposits are worth reviewing.

First, while this window was presented as a scheme to raise Non-Resident Indian (NRI) funds, only a fraction of the $26B was true NRI money. The rest was bank money coming in as NRI deposits.

Second, the concessional swap was only offered to banks and NRIs, allowing them to make some high returns at a cost to RBI. Are there ways in which such a scheme could be better priced while extending it to other deserving stakeholders as well?

The NRI scheme that wasn’t

Let’s understand how this scheme really operated for the NRI. Consider an NRI with $100,000 to deposit. The proposal her bank would have shown her in late 2013 would have looked somewhat like this (rates are illustrative):

Take your original $100,000

Borrow an additional $900,000 from the bank in Singapore for 3 years, at 2% p.a.

Place the entire $1,000,000 as a 3-year FCNR deposit in Mumbai at 3% p.a.

On redemption of the FCNR deposit, pay back the $900,000 loan, and take your net $100,000

Each year, you earn an interest of $30,000 (3% on $1,000,000)

Each year, you pay an interest of $18,000 (2% on $900,000)

On a net basis, each year, you earn a leveraged interest of $12,000 – or 12% in US$ terms - on your original $100,000!

Some banks even went a step further, and offered $1,900,000 of leverage against every $100,000 that the NRI put up.

The depositor had to contend with a cross-border transfer & convertibility (T&C) risk. What if the unthinkable happened, and India placed capital controls disallowing repayment of the FCNR deposit at maturity? The depositor would then need to repay her loan to the Singapore bank, without getting the funds back from the FCNR deposit. Did this (rather extreme) T&C risk justify the effective, leveraged double-digit rate of USD earnings to the depositor?

No. Overseas branches of Indian banks, foreign banks and other global investors did indeed take on India cross border T&C risk at far, far lower implied rates, even at that point in time. Additionally, if the swap window had been on offer against external commercial borrowings (ECB), then local companies (including the likes of IRFC/ REC/ PFC/ IIFCL/ EXIM) could have brought in US$, with the cost of T&C already embedded into the ECB loan pricing.

What was in it for the bankers?

The bankers – overseas and in India - managed a good deal as well. Again, rates below are illustrative.

The Singapore bank branch would have a low-risk loan out to the NRI at 2%, backed by the strong collateral of the FCNR(B) deposit. A typical foreign bank’s actual cost of 3-year funds at that point in time would have been about 1%, so they would have earned a 1% spread on a practically cash backed loan. 

The India bank branch would accept the FCNR(B) US$ deposit at 3%. Incremental FCNR deposits were exempt from the usual statutory reserve burdens of CRR and SLR. The bank would swap the principal with the RBI at the subsidized rate of 3.5%, and effectively raise 3-year Rupee funds at about 7.5%. This was at a time when the Indian government bond yield itself was above 8.5%. At the minimum, the bank raised 3-year money at 1% cheaper than true market, and therefore made a healthy spread on deployment. They could also pretend they had a balance sheet healthily funded by term client deposits - never mind that the deposits were largely funds from their own overseas branches! 

Overall though, there was a touch of irony in the trade for the bankers. After all, they effectively lent money to the NRI, only to borrow it back across the border at a 1% higher rate. That was the price they (and India) paid to access the RBI subsidized swap window, which was solely on offer against NRI deposits.

In Summary

Bulk of the money that came in the form of FCNR(B) deposits during September – November 2013 was not NRI money at all. It was overseas bank & institutional money that was funneled in as NRI deposits. 

For allowing us to showcase this – somewhat misleadingly - as diaspora money helping the country at a critical time, NRIs earned double-digit US$ returns on their true corpus, high in relation to the risk they undertook.

Both the overseas and local branches of banks participating in the scheme made good returns from the scheme as well.

While the net cost to India was not obvious, with 3-year INR GOI bonds at around 8.50%, RBI (and India) effectively raised 3-year US$ at about 5.00%. This was about 4.35% over average 3-year US Treasury yields of the time. This was a high price to pay, even at that time.

To be fair though, the actual cost paid overseas via the scheme was the 3% on offer against the FCNR deposits. The balance 2% was effectively paid by RBI/ India to local financial ecosystem.

Putting this in context

We must appreciate that the circumstances at the time were unique.

Our currency markets were in turmoil, and it was important to shore up confidence quickly. A direct Sovereign bond under such circumstances would have been risky. There simply wasn’t the time to debate endlessly in search of the perfect scheme.

From the RBI’s perspective, rather than a locked all-in cost, the window was a subsidized forward cover written by them. The realized cost, over time, would be a function of how markets behaved. Given the stability the scheme managed to bring into currency markets, RBI saw the scheme as a gamble that paid off.

Indeed, the swap windows were very successful in addressing their primary objectives – raising considerable amount of tenor funds at short notice, while arguably keeping the costs somewhat open. The scheme changed the narrative around our currency at a time when India was counted amongst the fragile five countries in the world. 

Having said that, with the luxury of time and hindsight now available, it is worth considering how future schemes could be designed should the need arise again.

Possible alternative structure – more equitable, more transparent, less costly

We have to assume that during the next contingency as well, a direct sovereign issuance would not be an option.

Perhaps an alternative way of structuring a subsidized swap window in future could be to offer it uniformly both against funds raised by banks overseas, and against funds raised by their customers through approved borrowing/ deposit routes.

This would mean that this subsidized swap window would not be restricted to FCNR(B) deposits alone. Banks would be allowed to tap the swap window against their own borrowing from overseas, and for swapping ECB funds raised by their clients as well. With the benefit of the swap now available to the likes of IRFC/ PFC/ IIFCL/ EXIM, this could also help channeling of funds into productive investments.

Besides being more transparent, the scheme could then be cheaper as well. In the FCNR(B) examples above, the NRI earned 1% per annum on the leveraged bank funds since this was exclusively an NRI window. Note that the 2013 RBI swap window for bank borrowings raised $8B with a swap subsidy of only 1%, against the 3% subsidy for the NRI scheme. In all, we could save 1-2% of cost for dropping the NRI exclusivity & pretense, and making the swap available to all legitimate term sources of foreign currency.

For the equivalent of the FCNR(B) scheme of 2013, that would translate to savings of $250-$500 million per annum – material by any yardstick.