Early in 2022, Bangladeshi worker Fakhrul borrowed 350,000 taka to fund his recruitment cost. More accurately, the agent’s fee was 300,000 taka, but Fakhrul must have felt that he needed a bit of extra cash on hand to tide him over to his first payday.

At that time, the exchange rate was about S$1.00 = 63 taka. The borrowed amount of 350,000 taka would have been roughly equivalent to S$5,500,

He borrowed “from NGO”, Fakhrul said. When we asked for the name of this organisation, he said “Brac”. And indeed, on this page of the Brac website, we see a statement “Loans for workers seeking employment abroad, complemented with services from BRAC’s migration programme…”

Fakhrul knows the numbers well. The deal was that he would have to repay 35,000 taka every month for 13 months. The total repayment would therefore be 454,000 taka over a principal of 350,000 taka. The difference would be a sort of service fee: 104,000 taka.

104,000 taka is 29.7% over the principal sum of 350,000 taka. One might be tempted to think of 29.7% as the interest rate, but this would not be correct. Firstly, we normally speak of interest rates on an annual basis. So, 29.7% over 13 months is more like 27.4% over 12 months. But, more crucially, a service fee is not interest at all; it completely understates what the effective interest rate is, as will be explained below.

After making four months of repayments, Fakhrul suffered a work accident and became unable to work while he recovered. With repayments suspended, he is in financial trouble. He is not even sure what his suspension of repayments will mean in terms of rescheduling and new (likely higher) monthly amounts whenever he can resume repayments.

### Alim repaid his loan

Unlike Fakhrul’s, Alim’s story is of a loan taken some years back when he first came to Singapore. He too borrowed from “an NGO”, but has forgotten the name of the organisation. He borrowed 400,000 taka to fund about two-thirds of his training and recruitment cost. The rest he secured through loans from relatives.

At that time, the exchange rate was roughly S$1.00 = 61 taka, which means that the 400,000 taka he borrowed was about S$6,500.

Alim’s story has a happier outcome. He managed to repay the loan in full. Although it varied somewhat month to month (“because some month salary up, some month salary down”), he repaid an average of 30,000 taka a month over a period slightly more than two years.

Using a multiple of 24 months (for simplicity) we estimate he repaid a total sum of 720,000 taka. This meant that the excess (service fee) over the principal amount was 320,000 taka – representing 80% over the principal sum of 400,000 taka.

Once again, one might be tempted to think of an 80% service fee over two years as an interest rate of 40% a year. But this too would badly understate the effective interest rate.

### Why not approach a bank?

Before we deep-dive into the numbers, we should place on record here Fakhrul’s and Alim’s answers to our question, “Why did you borrow from an NGO, why not from a bank?” Their replies were similar. Alim said “Bank ask for many things”, suggesting that they had more onerous requirements regarding collateral, guarantors, proof of income of guarantors and such like. Fakhrul also added that it would take much longer to arrange a loan from a bank.

### If converted into a conventional interest and repayment schedule

To see why we say the service fee in both Fakhrul’s and Alim’s cases understate the effective interest rate (in the conventional sense), it is important to bear in mind that if monthly repayment is required (as was the case for both men), then a portion of each monthly payment should go to reducing the outstanding principal amount.

The interest payment for the subsequent month should then be based on the reduced outstanding principal, and not on the original loan amount. This naturally makes for very complex calculations.

Below is a table showing two renditions of Fakhrul’s loan using this method.

Column C assumes a straight-line monthly repayment of the principal; he would have to repay 26,920 taka of the principal each month, over 13 months.

Column D uses the service fee percentage as if it were the interest rate. We had earlier shown that his service fee was 29.7%, spread over 13 months, or 27.4% per annum. But if we use this rate on the monthly-reducing principal, he would need only pay a total of 56,000 taka on top of repaying the principal instead of the 104,000 taka based on his agreed repayment schedule.

Column E shows what it would take to make his interest payments add up to 104,000 taka. It would take **an interest rate of 50.9%.** That is the true interest rate (in the conventional sense) behind Fakhrul’s loan.

The next table is a similar calculation for Alim.

Again, Column C assumes a straight-line repayment of his principal. Since it is spread over 24 months, it is 16,670 taka a month. Each repayment causes the outstanding principal amount to be reduced for the subsequent month (Column B).

We had earlier said that Alim’s service fee of 80% over two years looked as if the interest rate was 40% per year. But if we used that figure (in Column D) we find that his interest payable would only add up to 166,670 taka over two years, vastly less than what actually happened: he paid an excess of 320,000 above his principal.

In order to arrive at 320,000 taka as total interest paid (to match what Alim actually paid), we need **an interest rate of 76.8%** (Column E).

### Why so high?

Interest rates for bank loans vary from country to country, but fundamentally, they should bear a relationship with the inflation rate of the country. For the decade from 2012 to 2021, Bangladesh’s inflation rate held quite steady, at about six or seven percent per annum.

At first sight, it is shocking how high the interest rates were that our two men had to bear for their loans – 50.9% in Fakhrul’s case, 76.8% in Alim’s case. They are completely disproportionate to Bangladesh’s inflation rate.

One possible reason for this very high multiple over the inflation rate would be that the lenders judged migrant worker loans to be very risky, with a historically high failure rate. We don’t have the data, but it could well be true. A significant percentage of such borrowers might at some point default on their loans. Between our two examples, one of them (Fakhrul) suspended repayments after his fourth month.

But if indeed a high percentage of migrant worker borrowers fail to repay their loans according to schedule, does that not mean that for a high percentage of them, their migration careers are not going smoothly – commonly known as “unsuccessful migration”?