There is a great amount written and discussed about capital flight from Bangladesh.
The mechanism by which this happens is never made very clear.
Further, there is a great deal of discussion about the flow of remittances to Bangladesh.
There is an official picture of the balance of payments that is the basis for making economic policy.
Almost all of these discussions are wrong, distorting the picture of the economy as it actually functions and the purpose of this article is to explain the actual mechanisms and present a somewhat different picture of the balance of payments.
The hundi market
I describe the manner in which this market works by a simple example.
Mr XYZ wants to buy a house in Canada that costs the equivalent of $100,000.
He contacts one of the brokers of the Hundi market and explains his wish.
The broker tells him to provide Tk90 lakh. This includes the fees and everything that the broker will earn.
After handing over the Tk90 lakh, Mr XYZ finds $100,000 in his account in Singapore, which he transfers to Canada.
Where has this money come from?
The group participating in the Hundi market obtains dollars by buying them from Bangladeshis working in the Middle East.
The Bangladesh worker in KSA, Mr ABC has saved $5,000 and wants to send it to his mother in Rangpur.
He has two paths: He can take the money to a bank or finance company and send it to his mother.
The bank sends the money to the foreign exchange account of the bank in Bangladesh.
The bank calculates the Taka equivalent; if the exchange rate is Tk85 per dollar, this amounts to Tk4.25 lakh [We ignore the fees that the bank will charge.]
The bank informs the mother that the money can be collected at the branch of the bank nearest to the mother’s residence.
She then travels to the bank and goes through the formalities and receives the money.
The second path for Mr ABC is to go to the hundi market in Saudi Arabia.
The hundi market operator will obtain the address of the mother and will usually offer a better exchange rate say Tk86 per dollar and no fees.
Some of the Taka given to the Hundi broker by Mr XYZ will be delivered to the mother of Mr ABC.
Usually the hundi will deliver the money directly to the home of the mother who will not have to travel to the bank branch and go through all the paperwork to get her funds.
It should be obvious that by adjusting the exchange rate offered to the worker in Saudi Arabia the hundi system can attract whatever money it needs.
The government gives an extra 2.5% incentive to use the banking system; the hundi system just matches this.
The commercial bank in Bangladesh often does not treat low income old women very well and the mother may wait all day to get her funds.
At the end, Mr XYZ has his money in foreign currency; mothers have received their small amounts of Taka from what Mr XYZ has handed over and the difference in the exchange rates has earned the hundi system the money to operate.
With the amounts used in this example the hundi earns almost 5% on the transaction.
Notice that no money crosses the borders of Bangladesh.
There are no bank records of any transactions; Mr XYZ's foreign bank may ask where the money is coming from but probably not, unless the amount is very large.
The authorities in America or Canada or Singapore do not see an illegal transaction anywhere.
Everyone knows how this hundi market works.
It depends upon the Bangladeshi workers wanting to send money back to their family and finding the hundi channel gives them a better deal than the banking system.
As the hundi system has complete flexibility in setting the exchange rate it has the ability to adjust the terms according to the demands for dollars from persons who have Taka.
Another important aspect is that the hundi operators get the money to the recipients in a simpler, more pleasant way than do the banks.
It should be noted that mobile financial systems such as bKash have nothing to do with the hundi market.
The hundi market may use the mobile financial systems to move money inside Bangladesh to the recipient, which is better for all involved.
But now the MFS is a substitute for the delivery girl on a motorbike, not an inherent part of the hundi system.
We also note that there are many participants in the hundi market, it is efficient and competitive.
The hundi system has developed because the central bank in Bangladesh has put in place controls over capital flows.
If you live in Japan and want to send $100,000 to buy a house in Canada this is a routine transaction.
But in Bangladesh the central bank controls and usually forbids outgoing foreign exchange transactions other than paying for foreign goods and services.
There are recent changes in regulations to permit purchase of assets abroad in connection with Bangladesh industry, but capital controls remain very stringent.
Supply of foreign exchange to the hundi market
Foreign exchange going to the hundi market is virtually all from the earnings of Bangladeshis working abroad and remitting money back to Bangladesh.
The remitter has the choice as described above between the banking channel and the hundi channel.
The total volume of remittances depends on the number of workers abroad [including persons who have immigrated to the United States and want to send money to their family in Bangladesh], how much they are earning abroad and how much they want to send home or spend where they are living.
How this total volume of remittances splits between the hundi channel and the banking channel depends on the demands rising from the hundi market; what comes through the banking channel is really a residual.
The government and Bangladesh Bank are always claiming that they are taking actions that influence the split between the two channels.
But in reality there is little that they can do.
The hundi market sets its own price so can always adjust to increase the amount it takes off the remittance flows.
The government’s program to give a special exchange rate for the remittances passing through the banking channel accomplishes nothing.
There may be a short period where the hundi market receives less than they want, but this is corrected very quickly.
There are a number of surveys that attempt to determine how much of remittance funds received by Bangladesh households come through the banking system and how much through the hundi system.
The rule of thumb at Bangladesh Bank has usually been half through each channel.
The surveys show a range of splits averaging 55% through hundi channel and 45% through banking channel.
The range suggests the hundi channel will take 1.3 to 0.9 of total remittances.
For FY19, remittances through the banking channel came to $16.5 billion.
At the upper limit this implies $24 billion through the hundi channel.
This suggests about $20 billion came through the Hundi channel in that fiscal.
We conclude that the supply is in the range $17-23 billion.
Demand in the hundi market
There are a number of sources of demand for use of the facilities of the hundi market.
The most important is to support under invoicing of imports; the second greatest demand comes from foreign persons working in Bangladesh remitting part of their salaries home.
The substantial informal trade between India and Bangladesh, largely imports into Bangladesh, may be partly settled through the hundi market.
Finally, travel to India for education, medical treatment or tourism may be financed through the hundi market.
Capital flight is the final source of demand.
Under invoicing is widely practiced for imports from China and India; to a lesser extent from Japan particularly with respect to automobile imports.
This is motivated by the high import taxes [the total import charges include many different taxes: customs duties, supplementary duties, VAT, Advanced Income Tax.]
The total tax rates for imports of consumer goods are of the order of 40%.
Hence if the invoice cost of the imports is lower, the import tax is lower.
The Bangladesh importer pays for the import partly with a letter of credit and partly with a bank transfer.
The letter of credit is the official record of the import and authorizes the use of foreign exchange to pay for the import.
The bank transfer is handled through the hundi market.
This transfer is unknown to Bangladesh Bank and to customs.
The import data will reflect only the L/C amounts.
Hence the official record of imports underestimates the actual value and composition of imports.
This method of paying for imports is going on all the time.
During the period when customs employed pre-inspection companies, there was a great deal of evidence of under invoicing.
Unfortunately this has never been used to obtain a more accurate picture of the macroeconomy.
Indeed the few public statements have suggested that this is not important.
However, using pre-shipment inspection data the imports from China are 50% more than recorded; for India 30% more.
For 2018-19 the unrecorded imports from these two countries came to $9 billion; total under invoicing from all imports probably reached $13 billion.
When the PSI was in place NBR would charge taxes on the PSI company’s revision of the import value; however, no change was made in the value of imports.
The second use of the hundi system is for remittances of workers in Bangladesh to other South Asian countries.
These workers are largely in the textile and RMG sectors.
The salaries paid for these foreign workers should all come under the formal balance of payments.
For India for example there is virtually nothing in the Bangladesh Bank's balance of payments for remittances to India by workers in Bangladesh.
The total amounts of the remittances made by foreign workers are much discussed.
We estimate this is about $3 billion/year.
This is probably on the low side.
The informal trade across the Indian border is very difficult to estimate in total.
The existence and mechanisms of this are well known.
We estimate the trade as net $3 billion of imports into Bangladesh. The limited number of estimates of such trade is $3-4 billion a year.
Visitors from Bangladesh to India come in very large numbers for education, medical care, and tourism.
In 2019, some 2.6 million Bangladeshis visited India.
Most of this expenditure is financed through the hundi market. Although in this case the Bangladesh traveler wants Indian Rupee, not dollars. For 2018-19 we estimate this totaled $1.7 billion.
In total we have identified $21 billion demand in the hundi market in 2018/2019.
The range we estimate as $18-25 billion.
Forrest Cookson is an economist who has served as the first president of AmCham and has been a consultant for the Bangladesh Bureau of Statistics
