Muhammad
Grameen Bank

On subsidies, I've looked closely at Grameen's 2000 and 1997 borrowings statements and can't see how the soft loans have been reduced. The IFADs have been reduced. Payments haven't begun on 2% loans from Norway and Sweden. The Japanese 2% loan is bigger now than in 1997, and there is a new loan from Grameen Fund. So, by my calculations, 2% loans account for 52% of outstanding borrowings, compared with 47% in 1997. Even if Grameen is paying market rates on the new bonds, at least half the outstanding borrowings reflect effective subsidies, right? If we cite the Khalily study, should we say Grameen disagrees with its methodology/conclusions and doesn't think it has any subsidies? Should we say Grameen says it is moving away from subsidies?

2.1 I am attaching several tables showing the amounts outstanding against borrowings from abroad and local banks during 1997-2005. I have divided them into two categories, soft loans (defined as loans at interest below bank rate), and market-interest loans. (All loans with market-interest will be repaid by 2002.)

Please notice that soft loans declined in amount from Tk. 11.06 billion in 1997 to Tk. 7.24 billion in 2000. (I hope now you can see the point I was making.)

2.2 NORAD and SIDA loans that you refer to are loans given at perpetuity. So, they'll always be there.

2.3 Last trance of JBIC loan was received in 1998. This is the reason why JBIC loan was bigger in 1998 than in 1997. JBIC was ready to give us another loan, twice as much as the first loan. But we decided in 1995 that we are not going to take any more grants or soft-loans. So, we did not accept it. They were unhappy with our decision since they already had gone through their preliminary approval process on the second loan.

Let me also draw your attention to the fact GB gives housing loan at the subsidized rate of 8 per cent. JBIC loan was primarily for housing loans. This was a case of subsidized loan for GB to be used for subsidized loan for GB borrowers. When we discuss subsidy for GB we can also mention that GB has a large component of its outstanding loans devoted to subsidized loan programme given at a rate 60 per cent below the regular interest.

2.4 Grameen Fund loan has an interesting background. We had to accept the loan at the Finance Ministry's pressure. This was an international loan which government had accepted, but could not utilise. Government requested GB to take this loan from a government-owned bank. We came up with a proposal for the government. Grameen Fund will receive the loan from the government bank, Grameen Fund will lend the entire amount to Grameen Bank at 10 per cent interest. Government agreed. That's how this loan came to GB.

2.5 Please notice the 2% loans as percentage of total borrowing in Table 1(last line). It has steadily increased. By definition that's what will happen. In 2005 it becomes 100% of total borrowings ! The reason is we are paying off the loans whenever they fall due, and we are not taking any fresh loans (soft or full-cost). Ultimately total borrowings get reduced to only long-term soft loans.

2 per cent loan from IFAD will be entirely paid off in 2039. Only 2 per cent loan that will remain after that year will be the perpetual loans.

If we look at 2% loan as percentage of total borrowing it will give misleading conclusion. This percentage will become larger because denominator is becoming smaller faster than the numerator. I think the meaningful information can be produced by taking 2% loan as percentage of annual outstanding loans, or as percentage of total financing (borrowings + deposits + capital + reserve). Alternatively, looking at the absolute amount of 2% loan will immediately give the real picture.

2.6 We did have some comments on Khalily study. I'll try to dig up some of these comments later.

2.7 Yes, indeed, Grameen is moving away from subsidies. GB is not taking any new grants or loans (soft or full-cost) and is paying back old subsidized loans. Please see the Tables. (This year we excused ourselves from receiving a soft loan of $ 4.0 million offered to us by the Government of Spain at their own initiative.)

Is Grameen Bank requiring its employees to make up personally any unpaid balance of the contract loans? When employees retire, do their benefits depend on whether their borrowers achieved 98% loan repayment?

3.1 That would be a self-defeating (even suicidal) step to take. If I were a branch staff and I was told that I'll have to pay the unpaid installments of a loan that I gave, out of my own salary, I would not give this loan in the first place. I would do everything in my capacity to make excuses why I could not give that loan.

Secondly, the staff will rebel in a body if GB asks for that. Who would use up their small salary, on which their families depend, without protest, just to keep the job ? Practice in Bangladesh is to make extra money in bribes and other benefits when you get a job. People expect as if right to accepting bribes comes as a built-in privilege with a job. We are lucky that the Grameen staff put in hard work for their salary alone --- they have learned to live with no bribes, no illegal extra money. On top of it if they have to give up part of their salary to compensate for a borrower's failure to pay back her loan, it will be asking too much.

Even if by some magic they agree to do that how deep is their pocket ? How many loans can they finance from their own pocket ?

My answer to your question is : No, it is not correct.

3.2 No, that's not the rule. Anybody who puts in service in GB for ten years or more, gets retirement benefits in cash, and gets it pretty quickly. No other conditionality is attached with it. We are the speediest paymaster in Bangladesh in handing over pension money. In 90 per cent of the cases maximum time needed to get the pension cheque would be less than a month. We have instances where we handed over the money as fast as within 24 hours in cases of emergency.

If we put a condition that the staff must have "X" per cent of repayment record, what will happen ?

a) Nobody will work in a weak branch.
b) Which repayment record should we consider ? Repayment record of the entire life time of the staff ? Last year of his service ?
c) How do we keep records for each staff, each year ?
d) Staff will be enormously encouraged by this rule to manipulate repayment record so that he can get the retirement benefit.

Last year 1477 staff took retirement and got a total pension benefit of Tk 680 million (more than the amount we put under loan-loss provision !). Average pension benefit came to nearly half a million taka for each staff retired. A similar number will retire this year. Every one of them will get his retirement benefit. If we applied the crazy rule that you suggest ---- make a guess how of many of these retiring staff will get retirement benefits ?

Our retirement benefits are so attractive many of the staff retire just to get that benefit. That's a large amount of cash to have to start a life all over again. Sometimes they put the entire pension money in a fixed deposit account and earn a monthly income of over Tk 1100 from interest. This is almost one-third of their present salary ! This they find very attractive.

So far we have paid out Tk 1.4 billion in retirement benefits !

We understand that Grameen makes a loan-loss provision of 100% for loans 2 years overdue, as do PKSF borrowers. I see 623 million taka in provisions on the income line in 2000. Following PKSF guidelines of an additional 50% provision for 1-year overdue loans would have meant an additional provision in roughly the same amount, right? (1-year overdue loans being twice as much as 2-year overdue, and the percentage being half as high). So is it correct to say Grameen would have shown a loss of about 600 million taka in 2000 had it followed PKSF guidelines?

4.1 Yes, of course, you are correct. If GB doubles its loan-loss provisioning, GB would incur a large loss in 2000.

But the real issue is, why must GB double its loan-loss provision ?

GB's total expenditure in 2000 was Tk 3.0 billion including provision. Expenditure for provisioning was Tk 0.62 billion. This accounts for 21 per cent of the total expenditure for the year 2000.

If GB doubles the provisioning by another Tk 0.62 billion, the total cost will be Tk 3.62 billion. Provisioning as percentage of total expenditure will be 34 per cent.

Your way of posing the question only detracts us from the essence of the issue, that is, financial vulnerability of GB arising out of non-repayment of loans and interest by the borrowers.

One advantage with GB is, it is an institution with a long history. We can check with historical records. Purpose of provisioning is to make sure we do not show in our books some assets that we do not have, and not create a false feeling of having them.

By provisioning we wish to dispel from the minds of the owners of the company false sense of security and expectation of income. But it should not encourage us to create over-zealous provisioning by making provisions also for good loans.

Under-provisioning leads to exposure to risk. That's why it is not a prudent practice. Over-provisioning is also not a prudent practice. It adds to the cost unnecessarily. We should aim at optimal provisioning. While we should be keeping the organisation out of financial vulnerability, at the same time we should not get it involved in increasing the cost requiring an increase in price.

Under any scenario of non-repayment of loans, GB is well-equipped today, with its existing provisioning policy, to maintain its financial health. Why add further cost by making superfluous provisioning ? Why position ourselves to raise interest rate for no good reason ? Interest rate is already high.

I am insisting that GB is in sound financial health. It is adequately protected from financial vulnerability with the existing provisioning policy. Now, prove me wrong. That's the real challenge. Not whether GB is following somebody's grammar of provisioning or not. GB, by its very nature, has to be creative, not follow blindly something which has been designed in another world of banking. Our basis of banking is different, our clients are different, our relationships with borrowers are different, the premises on which we built our system is different.

Are we looking for a grammatical mistake, or we have a substantive cause for concern --- that's what we'll have to sort out. As I see it, there is absolutely no basis for concern. We are generously covered for possible eventualities of non-repayment.

Making 100 per cent provisioning for passed 52 weeks amounts puts too much pressure on the staff and the organisation. This may encourage the staff to seek artificial redress to overcome this pressure by wiping out "overdue" amount by issuing new loans. (In that case monitoring of number of "missed weekly installments" becomes very important.)

As long as we can take care of all potential bad debts without unnecessarily raising the cost of operation, we think that would be the optimum strategy. We find our procedure completely satisfactory from all points of view.

4.2 PKSF

You mention that PKSF's partner organizations follow a provisioning rule of 50% for 1-yr overdue loans.

You may have misunderstood PKSF. None of the large NGO's funded by PKSF (such, BRAC, ASA, Proshika) follow that rule. Their provisioning is done on the basis of disbursement. Following are their provisioning rules :

ASA 1 per cent of annual disbursement

BRAC 2 per cent of annual disbursement

Proshika 3 per cent of annual disbursement. Proshika raised this percentage from 2 to 3 per cent last year.

If you check BRAC's annual report of 1999 you'll notice that the following provisions were made for 1998 and 1999 :

1998 : Tk 250 million
1999 : Tk 220 million

Cumulative disbursement of GB from inception to 2000, stands at Tk 138.07 billion. Net provision (gross provision minus amount written off as bad debt) made during this period is Tk 3.78 billion. The percentage of loan-loss provision over the cumulative disbursement comes to 2.74 per cent.

BRAC's cumulative disbursement from inception to 1999 comes to Tk 41.0 billion. (source : Annual Report 1999). Net provision made during this period is Tk 0.47 billion. The percentage of loan-loss provision over the cumulative disbursement comes to 1.15 per cent.

GB has made Tk 0.62 billion loan-loss reserve in 2000 against the total disbursement Tk 14.04 billion in that year. It works out to be 4.45 per cent of disbursement. (Please see the Table for year-by-year percentages.)

No matter which way you look at it, our provisioning policy is quite robust and reliable.