Sri Lanka should stop surrendering Treasury dollar inflows to the Central Bank: Bellwether
COLOMBO (EconomyNext) – Sri Lanka should immediately halt the practice of surrendering government foreign loan proceeds to the Central Bank which will reduce pressure on the rupee and also allow corrective interest rates to be lower than it would otherwise be.
Why Sri Lanka is more prone to BOP problems than other country, and corrective interest rates have to go up higher than other countries (like India, Indonesia, Vietnam and of course China) is partly due to the existing procedural flaw, of surrendering state loan proceeds to the Central Bank.
As long government dollar loans are surrendered to the Central Bank there is no true float or market determining of the exchange rate.
As a result there has to be an un-necessarily hard landing in the economy to correct any credit bubble or ‘overheating’
Sri Lanka has had a practice of surrendering inflows to the Treasury to the Central Bank for decades. This has been seen to be done even during balance of payments crises earlier.
In 2011 July/August for example, dollars from a one billion dollar sovereign bond was converted in a day without even one dollar hitting the forex market, and a great many billions of rupee liquidity was instead created in one day.
The last BOP crisis could have been averted to a great extent if this was not done.
The sharp increase in excess liquidity (see graph August 2011) shows that dollars have been converted into rupees in one go. This happens not just to proceeds of sovereign bonds but to any other loan dollars flowing into the Treasury, as well.
Treasury is for the most part a net borrower but its inflows hit the banking system as rupees after conversion at the Central Bank and not as dollars. As a result they contribute only to import pressure on the exchange rate.
It is exporter dollars and remittances that are always depended on to support the rupee according to central bankers, policy makers and other commentators. The new central bank governor has made similar statements.
This perception is absurd, when the Treasury is usually a net receiver of dollars. It must be kept in mind however than any inflow, including remittances and export earnings will eventually create an equivalent import demand, unless the rupee proceeds are mopped up permanently.
Having said that however, there is great merit in allowing the forex markets to work without weighing the odds against the rupees stability.
The current demand for dollars comes from the spending of rupees existing in the economy already.
If the Central Bank does not monetize debt or inject money through the reverse repo window (printing money by acquiring Treasury bills), this money has essentially come from dollars bought by the Central Bank in the past to keep the rupee pegged at a fixed rate and prevent its appreciation. In other words it is a result of historical unsterilized dollar purchases.
The point to note is that every dollar collected by the Central Bank as reserves – whether surrendered or bought in the market, is tied to a rupee – which will be demanded back eventually, if the liquidity is allowed to remain unsterilized.
Dollars from exporters or expatriate workers that are sold by commercial banks in the forex markets are completely fresh, unencumbered with any rupees already in the banking system. That is why they are able to lift to the currency, at least in the first instance.
Any surrendered dollars in Central Bank reserves cannot give the same lift to the currency because they will have to fight the downward pressure coming from its own rupees created a few days or weeks back.
Excess liquidity and coconuts
Even before January 29 budget with additional spending – the rupees from past dollar purchases, which were not permanently sterilized – were being loaned out, either to the state or to private parties and it was creating pressure on the exchange rate.
If the banking system loans out rupees in excess of inflows through deposits and loan repayments, the extra outflows are likely to hit the balance of payments as additional demand since the economy cannot adjust with additional supplies immediately. Imports are the only remedy to keep prices from spiking.
Economic students are usually taught inflation through a simple model. If there are 10 rupees and two coconuts in the economy, a coconut will be priced at 5.00 rupees each. If an additional 5 rupees are printed, coconut prices will go up to 7.50 each.
But that is not what happens in practice. If coconuts were a traded good, another coconut will be imported from abroad. In other words, the additional money created will hit the BOP.
That is because while it is one thing to imagine that an economy is self-sufficient as a theoretical model, in practice they are not autarkies. Therefore the IS/LM model taught to economics students is also flawed.
In the case of loaning out unsterilized rupees from the distant past, the effect on the bop is exactly the same as if the money was printed though T-Bill purchases.
The underlying problem in the credit system cannot be corrected fully by putting Treasury rupees to the market, but it will make the corrective measures less stringent and prevent the very hard landings that we see in Sri Lanka usually after a credit bubble.
The underlying problem can only be corrected by:
(a) a rise in interest rates to curb consumption and drive more deposits to the banking system
(b) a rise in interest rates to reduce private credit demand
(c ) the issue Treasury bills at higher rates to directly increase the volume of saved resources within the broader economy available to the budget.
(d) an increase in steady Government revenues to reduce its borrowing requirement or cutting spending such as civil service reform.
(a) a permanent cut is state spending such as trimming the civil service
(d) any sustainable private or state net foreign borrowing that goes on month after month.
A rise in interest rate can be implemented through policy rates and also by completely sterilizing excess liquidity, which will also curb credit and save foreign reserves.
Even here, by hitting dollars at the forex market slowly (say a billion dollars from a sovereign bond are put to the market 300 dollars a month or something like that) it is possible to buy time and make the adjustment less of a shock to the banking system.
However during a time of currency pressure, if the dollars coming into the Treasury are surrendered to the central bank, the effect in the BOP becomes worse than the underlying structural mis-match in credit and interest rates.
Any dollars surrendered to the Central Bank are taken away from the foreign exchange market, depriving any positive benefit. That means Treasury dollars are neutralized before they begin to hit the market via any Central Bank dollar interventions later, through the rupees that are created.
Central Bank dollar sales could give a positive lift only if there was an aggressive sterilized purchases done through higher rates.
If dollar sales are unsterilized – like now – they also do not cause any new harm, but they simply mop up the excess liquidity, and prevent any future import demand at the cost of forex reserves.
Eventually when liquidity runs out, unsterilized sales will end and instead sterilized sales will be begin. That means the Central Bank will actively contribute to pushing the rupee down, in a never ending vicious cycle, generating new import demand and a full-blown BOP crisis would have started which can only be corrected by a true float.
Before that happens, the dollar inflows to the Treasury has to be put into the forex market.
The very worst thing that happens in Sri Lanka is that Treasury dollars are surrendered to the central bank even when sterilized foreign exchange sales are made.
One option is to create a Treasury forex trading desk for the excess dollars which are not used for loan repayments. To minimize conflicts and future problems, activities of the trading desk can be limited to the only the sales of dollars.
The economic planning minister can set up such a unit and hand it over to the Treasury later.
The US had a Treasury trading desk to sell.
But may be a better and infinitely faster policy will be to farm out Treasury dollars to several banks to sell in the market and get the rupees from importers.
Since People’s Bank is always short of dollars part of the money can be given to them. The rest can be given to Bank of Ceylon.
Another model would be to give 50 percent to state banks and 50 percent to private banks, based on their trading volumes last year.
In the meantime, monetary policy has to be tightened to fix the underlying problem.
Difference between Treasury and Central Bank dollar sales
Unlike central bank interventions, Treasury dollar interventions are monetary policy neutral.
That means, while Treasury dollar sales cannot do any good in the long term, they can do no harm either. They do a great deal of good in the short term. They act like export earnings or remittances. They can prevent panic among importers. They can persuade exporters sell dollars instead of holding on as the pressure on the currency is removed.
Central Bank interventions on the other hand fundamentally affects the rupee because its actions are monetary policy positive, that is to say it affects the volume of the monetary base, and therefore the interest rate and the volume of bank credit.
A central bank will engage in four types of dollar interventions.
a) Unsterilized purchases – dollars are bought from the market or the Treasury and rupee are created, expanding reserve money and forex reserves. The rupee is prevented from appreciating.
b) Unsterilized sales – dollars are sold in the market, reserve money is shrunk as rupees are mopped up and the rupee is prevented from weakening.
c) Sterilized purchases – dollars are bought and rupees are mopped up. The rupee is pushed upwards because there will be no import demand from the purchased rupees.
d) Sterilized sales – dollars are sold and more rupees are created to re-fill the rupees that are mopped up by the sales, preventing the monetary base and credit from shrinking and the rates from going up. This is BOP crisis in action.
If the government buys dollars in forex reserves by selling T-bills to the central bank instead of rupees, that is also monetary policy neutral. Therefore foreign loans can be settled in this way, without affecting the exchange rate or domestic rates but a loss of reserves.
IMF transactions are also monetary policy neutral.
To conserve the existing forex reserves the existing excess rupee liquidity can to be sterilized for six months or more through the sale of Central Bank Securities or borrowed Treasury bills as soon as term auctions mature.
This will pre-empt the private sector from using the credit for imports.
Then banks will not be able to use that money for lending as credit demand go up and they will be forced to pay higher rates for fixed deposits and push up overall savings in the economy.
This will reduce consumption and reduce overall credit demand but it will cause market interest rates to move up faster. If reserve conservation is not a priority the excess liquidity can be loaned to the government (instead of new savings) and the foreign reserves run down faster.
If the excess liquidity was properly sterilized in 2014, when credit demand was low or negative, this problem would not have occurred.
But there is no point in crying over spilt milk.
Above all if the sovereign bond is sold, its proceeds should not be surrendered to the Central Bank. In the meantime apply the same principle to any net dollars from Asian Development or World Bank.
The interbank forex markets are small. Even an additional 10 to 20 million dollars of sales a day can make a big difference to the exchange rate.
This column is based on The Price Signal by Bellwether published in the March 2015 issue of the Echelon Magazine. To read Bellwether columns as soon as they are published, subscribe to Echelon Magazine at this link. The i-tunes app can be downloaded from here.