ECONOMYNEXT – Sri Lanka is now knocking at the door of a balance of payments crisis and for all intents and purposes seems to have passed the point where the problem could have been solved with a 100 basis point rate hike.
Rupee bond holders who were remarkably patient (or reckless) have started to sell out in earnest. That is why this column warned several times that Sri Lanka can no longer take the fiscal risks of the past, because we are now exposed to high levels of foreign market debt.
When foreign investors are spooked and they start exiting, the problem becomes much more difficult to solve. Sri Lanka has set in motion a beautiful pro-cyclical combination of fiscal profligacy and accommodative monetary policy to get to this point.
Let’s look at how a frontier market or third world central bank creates a balance of payments crisis. The 2011/2012 crisis, coming in the wake of a 2008/2009 crisis is a good starting point. The 2011/2012 crisis was triggered by credit-financed fuel subsidies, which were ultimately accommodated by the Central Bank. The roots of the crisis lay in a 2011 budget, when a decision was made presumably by the Finance and Power Ministries not to adjust fuel prices that year. The resulting downturn and currency depreciation, helped oust the Rajapaksa regime.
But bad fiscal decisions can be countered by tighter monetary policy.
The Root of All Evil
Currency pressure and inflation is created by domestic operations of a Central Bank, which is at the heart of monetary policy operations (the inflation generator) of any such agency. It is through liquidity injections (printing money) that all evils are let loose.
A few pre-conditions are needed to trigger a balance of payments crisis. Credit has to be positive and the economy has to be ticking along well. That is why countries that were doing well in East Asia fell prey to currency crises easily.
If there is a ‘liquidity trap’ and credit is negativeit is possible for the state to borrow or print and spend large sums without creating a crisis. A Keynesian stimulus if you will. Keynes erred in naming his theory a general theory. It should have been named a special theory applicable only to liquidity traps that is when credit is negative. If credit is ticking along nicely, and deficit spending starts, rates have to go up to curb much more productive private spending to maintain economic stability.
If that does not happen, a currency crisis is the result in a country with a pegged exchange rate.
Neo-Mercantilist theory taught to students as ‘economics’ including IS-LM are therefore completely false. Nobody lives in an autarky without international trade. When money is printed it will spill over to the balance of payments. ISLM-BOP or the Mundell-Fleming model is closer to reality.
Lets rewind a few years to understand the stage in the cycle we are now in.
Sri Lanka had a balance of payments crisis in 2009. In May 2009 the rupee was floated to stop the crisis. Until then money was printed by buying up Treasury bills. It was necessary to buy Treasury bills to ‘sterilize’ or neutralize the effects of a foreign exchange sale, which sucks rupees out of the system into the Central Bank.This is called a sterilized foreign exchange sale. When sterilized foreign exchange sales start, it is very difficult to pull back from a BOP crisis, without a free float and a rate hike.
In May 2009 the rupee was floated, and it was possible to stop sterilized foreign exchange sales. When interventions stop, liquidity shortages stop and then there is no longer a need to print money to fill the gap. After May there was a period when dollars were bought, creating liquidity and Treasury bills in the Central Bank’s portfolio were sold to sterilize or pull out the liquidity in a complete reversal of the earlier process.
That is a called a sterilized foreign exchange purchase. When such sterilized purchases of forex are made, dollar reserves go up, but the rupee liquidity that is generated is also extinguished and cannot be spent by anyone. As a result foreign exchange remains locked in forex reserves without being redeemed.
In 2010, the Central Bank bought back dollars and at the same time also allowed the rupee to appreciate back to earlier levels. In late 2011, the Central Bank stopped sterilizing purchases and allowed liquidity to build up.
Liquidity went up and down in some months but no sterilizations was done. That is a period of non-sterilized foreign exchange interventions. There were non-sterilized sales and non-sterilized purchases, and liquidity went up and down. That means the balance of payments and the economy is in sync.
As can be seen in the graph on domestic operations, from the beginning of2011 the central bank made steady unsterilized sales. The proximate cause was a drought, which was drawing more oil imports for power generation.
Accommodating Supply Shocks
Authorities could have either raised power prices (to neutralize the shock to the monetary system) or raised interest rates and refused to accommodate the negative supply shock that was drawing more credit from state banks to cover power sector losses. If more deposits could be raised the economy would re-balance.
On January 11, 2011 the Central Bank cut rates, accommodating the shock. That was the unkindest cut of all. It also continued to sell dollars. Dollar sales however draw out liquidity and it is an automatic adjusting mechanism in a pegged exchange system. This is a non-sterilized foreign exchange sale.Eventually when all liquidity dries up, rates should go up, nipping the problem in the bud.
When non-sterilized sales became persistent and one-sided in 2011 independent analysts warned at the time that Sri Lanka was heading for a BOP crisis. The Central Bank, in true emerging market style (read third world) delayed to act.After August 2011, when liquidity ran out, it started to buy Treasury bills and print money, when forex sales extinguished rupees. In other words it started a vicious cycle of
sterilized foreign exchange sales. This is classic balance of payments crisis territory.
Laster it made several failed floats. Though it floated in February 2012, it continued to intervene ostensibly to pay for oil bills. The most dangerous action a Central Bank can take is to make interventions in fits and starts and fully sterilize them. While it may be possible to hold a currency by fully defending and sterilizing partially, it is not possible to do the opposite. If interventions are made from time to time and sterilizations are full, there is no end to where a currency can fall to. That is why the rupee fell from February to May 2012.
A full float on the other hand means there will be no interventions. Consequently there is no need for sterilization either. So the vicious cycle of interventions and sterilizations stop.
After May 2012 the Central Bank started to purchase dollars and sterilize them again building up foreign reserves.
This time it’s different
But somewhere in 2014 the Central Bank came up with a new tactic that contributed to making the monetary system more prone to balance of payment crises and reserve losses. Instead of selling Treasury bills outright to kill liquidity, it partially sterilized liquidity through ‘term repurchase’ contracts.
The evil of the term repo contract is that, like overnight deposits, liquidity become available for banks to lend whenever a term repo contract ends.
This column late in 2014 warned that the strategy of sterilizing temporary was dangerous and when credit picked up and up to 2.4 billion US dollars could be lost very easily. At the time there were 8.8 billion dollars of reserves.
Term repos were used perhaps because some of the Treasury bills were borrowed from the Employees Provident Fund. The problem became acute partly due to steady transfers of profits to the Central Bank. These funds should have been kept to repay the IMF. When profits are transferred to the Treasury as rupee liquidity, either dollars are lost when they are spent, or more bills have to be borrowed to keep the liquidity. Either way it is a no-win situation.
The founders of the Central Bank had devised Central Bank Securities to solve the problem, but their use was abandoned several years ago. To compound the problem some Treasury bills, which were owned by the Central Bank were also used for term repos.
The IMF also perhaps did not realize that there was a credit risk to their loan in allowing 350 billion rupees of excess liquidity to remain in the system after domestic credit turned positive in late 2014. This was perhaps the reason for their highly optimistic forex reserve projections. In late 2014, there were also some selling by foreign bond holders. But that shock was easily absorbed by the pile of liquidity.
By then domestic private credit was picking up. The cycle had turned. Then liquidity and forex reserves started to fall. There was ample time to take corrective action.
In January there came another shock to the credit system in the form of a revised 2015 budget. The last administration was also in the habit of taking on increasingly higher snorts of international sovereign bonds to fix the deficit without putting too much pressure on the domestic credit system.
It is however not possible to carry on such a strategy forever unless there was space for real private sector growth. But due to a fixation with the state expansion, non-privatization, corruption, expropriation and lack of rule of law, many freedoms were curtailed and there was insufficient real economic growth that was generating tax revenues for the state.
The new administration then delivered several shocks to the system. State salaries and pensions were raised, and the planned 1.5billion US dollar sovereign bond was cancelled. Taxes were cut in several foods, which was good for tax equity. Fuel prices were lowered.
Subsidies were not only dished out willy-nilly at the expense of the poor, putting Mahinda Rajapaksa to shame, but even people’s freedom to buy drugs and the choice was to be controlled.
The Yahapalanaya shock almost like the failed ‘Rata Perata’ shock in 2004 and the Mahinda Chinthanaya that followed, came complete with anti-business rhetoric, but with less nationalist hate.In fact some UNP voters are now in a peculiar position where their representatives seem to have ‘crossed over’ to the Rajapaksa policy camp after taking their votes sans nationalism. Such is the policy fright of the UNP leadership.
When current spending was ratcheted up in January and petroleum prices were cut making it more difficult for state energy enterprises to repay debt, negative credit shocks were delivered to the banking system.When the sovereign bond sale was abandoned, yet another potential positive shock to the credit system was removed.All this demanded a rate hike.
On the surface however the inflation index fell, though beneath the surface the credit system was simmering.With private sector credit picking up, the government was no longer able to borrow the 2014 year’s credit volumes by at the same interest rate structure.
Authorities however took comfort in a falling inflation index, which was helped by tax cuts and also low external commodity prices coming from a stronger dollar. Even when oil prices picked up, they were not passed on, while salary and other subsidy costs also rose.
This problem of juggling with dual anchors is a common mistake that is made by third world central banks, helping them stay in the third world for a long time. The entire Bretton Woods system collapsed due to this problem of mistaken ideology.
If you are targeting an exchange rate, then the monetary system is anchored to that exchange rate and to that currency peg, however it might be otherwise wished. It is not possible therefore to target a domestic anchor (the consumer price index) at the same time and cut rates or print money. Then balance of payments trouble is inevitable. When interest rates and exchange rates are targeted at the same time, the currency collapses.
It is therefore necessary to radically reform the Central Bank. The other option is to abolish it and re-create a currency board, the arrangement that existed prior to 1951. If a central bank is to be prevented from de-stabilizing the economy every three or four years it should be made to target one anchor; a domestic one with a floating exchange rate or an external one with floating interest rates.
Throughout the current episode of BOP pressure, the Central Bank was releasing liquidity temporarily tied up through term repo deals. This was a type of quantity easing. Forex reserves were steadily lost as the released liquidity was used by the government or private parties which then generated imports.
If we define sterilized foreign exchange sales as the defining feature of a balance of payments crisis, then this period is not really BOP crisis territory, since the liquidity releases come ahead of the dollar defence. In theory the Central Bank can halt this at any time and prevent further pressure. But when liquidity was steadily released a rate rise was delayed in the market.
The effect on the economy was the same as a sterilized forex sale: interest rates were suppressed and foreign reserves were lost.
The April rate cut came on top of all this. This rate cut was also symbolically similar to the January 2011 cut. But liquidity releases were anyway doing the same thing earlier.
There are all kinds of ways reserve adequacy is measured. Many of them do not make sense. Talking about reserve adequacy in the absence of an interest rate is useless. At a given willingness of the Central Bank tokeep interest rates low and to sterilize forex sales, unlimited amounts of foreign reserves can be lost.
But a few things are different today compared to 2009.There is a debt to IMF which has to be repaid. This is not the same as central government or private debt. It is one thing to sell reserves to the government to settle foreign loans. These reserves can be recouped by restricting domestic credit. That is by selling down the Treasury bills taken by the central bank when reserves are appropriated.
But IMF loans are a direct liability of the Central Bank and they cannot be recouped by restricting domestic credit. Dollars coming from central bank swaps are also absolute borrowed reserves.
Ignoring other debt, it is also possible to simply measure forex reserves very narrowly as a share of the domestic money outstanding, which is the ‘reserve money’ or the monetary base of a country. That is the main liability of the Central Bank to the domestic economy.
But any partially sterilized excess liquidity also becomes a liability against foreign reserves when the contracts mature. So such volumes are similar to the monetary base in all respects, until they are sterilized permanently.Hence the warning by this column about excess liquidity in the past several months.
It can be seen from the reserve adequacy graph that the forex reserve position is now much lower measured against the rupees outstanding than it was in January 2012 shortly before the rupee was floated.In January 2012 reserves were 181 percent of the monetary base. There was no excess liquidity. Put in another way, the rupee liquidity that could come up for redemption was only 55 percent of forex reserves.
But in October 2014, reserves were only 122 percent of the total liquidity. It means after exchanging dollars for all rupees, 22 percent would remain. Expressed in another way liquidity was 81 percent of the foreign reserves. That is why this column was urging that the liquidity be permanently extinguished with outright sales of Treasury bills and central bank securities.
By May 2015 reserves were enough to cover 132 percent of liquidity. However forex reserves also contain a 500 million dollar swap from the Reserve Bank of India. Without it reserves were at the same level as in 2014 at 123 percent of liquidity. That is much lower than the 181 percent figure in January 2012. Or liquidity was 81 percent of forex reserves, whichever way it is expressed.
The problem of soft dollar pegs, and dual anchors is also expressed as the impossible trinity of monetary policy objectives. The impossible trinity says that it is not possible to have free movement of capital and independent monetary policy and also a fixed exchange rate.
But excess liquidity and sterilized foreign exchange sales can help trigger a BOP crisis purely from current transactions. The presence of excess liquidity can also make it difficult to recoup lost reserves that are used to settle foreign loans under conditions seen in Sri Lanka.
Sri Lanka also has a problem in that dollar flows to the State are surrendered to the Central Bank undermining floats.
However if rupee bond holders are spooked, more pressure comes. In 2012, the then administration was urged by the IMF to take action before rupee bond holders were spooked. The majority of them stayed remarkably calm until the situation was fixed by hiking oil prices and interest rates.
This time however they are selling out in some volumes. Almost like they sold in 2008 and 2009, when their total holdings were only 600 million dollars. Dollar bond holders are protected from currency depreciation.
On June 01, the central banks Treasury bill stock was down to 5 billion rupees. There was no more unexpired term repos, but excess liquidity was 97 billion rupees.
When such large volumes of excess liquidity is present, the rupee cannot be floated. The float will not take hold. Whenever the excess liquidity is used up, it will pressure the rupee further, similar to sterilized forex sales with partial defence of the currency. That is what happened in the period from February to May 2012.
So any advice to have a’flexible exchange rate’, when there is excess liquidity present is just nonsense. But whether depreciation can make foreign investors collectively refrain from selling due to continued losses is another matter. If the currency was allowed to appreciate back in 2014 there would have been more credibility of the peg or faith that after a float the peg will get back to earlier levels. It happened in 2009 with the rupee allowed to appreciate after falling to 119. And that may have been a reason for the rupee bond holders staying put in 2011.
Meanwhile the Treasury bill stock of the Central Bank has started to climb up. By June 18 it was up to 25 billion rupees from 5 billion at the beginning of the month, presumably due to foreign reserve appropriations to settle central government loans. By June 24 excess liquidity was down to 59 billion rupees from 97 billion rupees at the beginning of the month. A cycle of sterilized foreign exchange sales may be 400 million dollars away, until corrective action is taken quickly. Some forex borrowings can delay this. But it will not solve the problem.
By giving Central Bank swaps India has made the problem worse. The junkie was given another fix instead of rehabilitation. Swaps will not prevent sterilized foreign exchange sales either.
An IMF program on the other hand prescribes medicine to the central bank and the deceptive elected ruling class and withholds the release of foreign reserves, until medicine is taken. The rulers are then forced to tell the truth to the people, much to the anger of the deceptive elected ruling class.
The current rates are incompatible with domestic credit conditions. The deadly cut in January 2011 was to 7.25 percent. But the deadly cut in April 2015 was to 6.00 percent. Sri Lanka is that much more vulnerable.
This column is based on ‘The Price Signal by Bellwether‘ published in the July 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.