Lessons · Lesson 3 of 6
What a state export agency is actually for
Value a state export guarantee at the borrowing it unlocks rather than at the claim it might pay, and see why the correct expected-loss calculation gives the wrong answer.
Lesson 3 of 6 · 18 min
The agency is not selling you insurance
Almost every exporting country has one: a state or state-backed body that stands behind export trade. The names vary and the shapes vary, but the family is real, and it is called an export credit agency. Tamarask's is the Tamarask Export Guarantee Fund. This lesson is about what it is for, which is not what its brochure says and not what most finance managers assume.
The assumption is that it sells cheap insurance. You pay a premium, and if the buyer does not pay, the fund does. That is a real product, and course 13.5 prices it properly: what an insurer covers, what quietly voids a claim, and how a premium compares with a loss. Read that course for the economics of being insured.
This lesson is about the other thing an agency does. It does not appear in 13.5 because it is not a property of insurance at all. An export credit agency exists to change what your bank is willing to do. You are not really buying cover. You are buying a different bank.
What Ardhen offers once the fund is behind it
The Tamarask Export Guarantee Fund's pre-shipment guarantee covers 80% of a bank's advance against a confirmed export order. The premium is 0.95% of the guaranteed amount for the tenor, which is the length of time the guarantee runs.
Ardhen Bank's terms change like this:
| Without | With the guarantee | |
|---|---|---|
| Facility limit | USD 900,000 | USD 2,600,000 |
| Advance against order value | 60% | 85% |
| Interest a year | 14.5% | 11.75% |
| Blocked cash margin | 10% of the limit | waived |
Every one of those movements has the same cause. Ardhen is not being generous, and it has not changed its view of Semarra. Four fifths of the exposure now sits on the state instead of on Ardhen's own book. So Ardhen needs less collateral, prices less risk, and can write a bigger line without using up its own appetite.
The calculation the finance manager did, which was correct
Semarra's finance manager priced the guarantee twice, and both times honestly.
First, against the claim. Semarra's credit insurer rates Callowmere Group at 0.55% a year. On an order of USD 403,200 with 80% cover, the expected loss is USD 1,774.08. The guaranteed amount on this order is 80% of the 85% advance, which is USD 274,176. So the premium is USD 2,604.67. The premium is larger than the expected loss. On that test the guarantee is poor value, and the arithmetic is right.
Second, against the order's own cash cost. Semarra funds whatever the bank does not. Its cheapest alternative is stretching its trim supplier, which charges 2.5% for sixty days, or 15.0% a year. So the true cost of money on the order is a blend. Without the guarantee it is 60% at 14.5% and 40% at 15.0%. With it, 85% at 11.75% and 15% at 15.0%.
Run that over the order's own cash profile and the guarantee saves USD 1,614.08. Add the released cash margin, worth USD 807.28 an order from lesson 1, and the total saving is USD 2,421.36 against a premium of USD 2,604.67.
The guarantee loses by USD 183.31 an order. Twice measured, twice negative. The finance manager declined it, wrote a clear note explaining why, and was thanked.
The question neither calculation asked
Both tests priced the guarantee on the orders Semarra already has. Neither asked what it does to the number of orders Semarra can have.
From lesson 1, the Ardhen limit permitted 11.71 orders a year. Rerun those four lines with the guarantee in place. The consumption per order rises, because 85% of USD 403,200 is USD 342,720. But the limit nearly triples:
- Orders drawn at once: 2,600,000 ÷ 342,720 = 7.59
- Cycles a year: 3.15
- Orders a year: 23.87
Now the honest part, because this is exactly where a finance case turns into a fantasy. Semarra cannot sew 23.87 orders. Its floor, running the shifts it runs, makes 15.0 orders a year and no more. So the guarantee does not create 23.87 orders' worth of anything.
What it does is move the constraint. Before the guarantee, the binding constraint was the bank at 11.71 orders while the floor could do 15.0. So 3.29 orders a year of sewing capacity sat idle because of a credit limit. At USD 60,960 of margin an order, that is USD 200,807 a year.
The premium, at Semarra's old run rate, is USD 30,490.09 a year.
The guarantee returns 6.59 times its cost, and neither of the two correct calculations could see it. Both were per order, and the value is annual and structural. This is the lesson's whole point, and it generalises: an instrument that changes a limit cannot be valued on a transaction.
Moving the constraint from the bank to the sewing floor is the right direction of travel on its own terms too. A factory limited by its floor can fix that with a shift, a line balance or a machine. A factory limited by its bank can do nothing about it from inside the building.
What the fund will not do, stated plainly
An export credit agency is a state body with a mandate and a balance sheet. The honest limits matter as much as the benefit.
- The cover is not complete. Eighty per cent means twenty per cent stays with you, by design, so that you keep caring who you sell to.
- It does not cover a dispute. If Callowmere withholds payment because it says the shirts are wrong, the fund is not the answer. That is a commercial argument, and 13.5 explains why it cancels most cover.
- Claims are paid after a waiting period, not on the due date. The fund pays six months after the loss is established. A guarantee is not liquidity.
- The fund keeps a limit on each buyer and a list of markets it has closed. A limit can be reduced on goods you have already cut, and that reduction arrives as a letter, not a negotiation.
- The fund's paper carries the country's own standing. Wilmerton Bank, in Callowmere's market, will take a Tamarask Export Guarantee Fund undertaking, but at a haircut. It would take a Belveny one at par. A state guarantee is not a passport out of your own address. It is a discount on it.
What you should be able to do now
Ask your bank a question it is rarely asked: what would you do differently if this order were guaranteed by our export credit agency? Not "would you lend more". Ask for the specific movements. Advance rate. Rate. Cash margin. Limit.
Then value the answer in the right unit. Price movements are per order. Limit movements are per year, and they are worth the margin on the orders your floor could make and your bank would not fund. If your floor already runs full, the limit is worth nothing this year, and you should say so. If it does not, the difference between the two ceilings is the number, and it is usually an order of magnitude larger than the premium.