This whole certification teaches how to stand on the insured's side. This module changes seats and looks at the file from the chair opposite, not out of curiosity but because a professional who knows what causes a decline presents differently what can be accepted. A political risk underwriter does not compare a risk with a tariff: it assembles a judgment from elements whose order is stable and most of which appear nowhere in a questionnaire.
The first sort is not about the file, and that is what disconcerts an insured that comes to defend its own. The underwriter looks first at its own portfolio: what it already carries in that country, in that sector, and on that family of perils. An excellent file in a country where it has already committed its aggregation limit is declined with nothing held against it, and a mediocre file in a country where it holds nothing may pass. That stage is invisible from outside and it decides more often than the rest, which explains why the same file receives three different answers from three markets on the same day.
Then comes what the underwriter really looks for in the file, and it is not a probability. It looks for what it will be able to do if the event occurs: is there a bilateral treaty opening a remedy, does the ownership structure let it be used, is a multilateral carrier in the syndicate, does the asset have a value a market can recognise. It is the subrogation reasoning applied before the loss: an exposure from which nothing will ever be recovered is rated as an outright loss, and an exposure backed by a workable remedy is rated as an advance of cash.
The third element is the one insureds most underestimate and it bears on them rather than on the country. An underwriter assesses the insured: the quality of its local governance, the identity of its partner, its conduct in earlier claims, the way it answered the questionnaire, the depth of what it disclosed unprompted. A file containing an unfavourable fact volunteered is worth more than a smooth one, because the first informs about the person as much as about the risk, and it is the person who will produce the documents on the day.
One must then name what causes a decline rather than a higher rate, because those are two different decisions and not two degrees of one. A risk one understands and can bound is rated higher. A risk one cannot bound, cannot model for want of any comparable, whose correlation with the portfolio is unknown, or whose settlement would be prohibited by a sanctions regime, is declined. Also declined, and this is said more rarely, is a file where one suspects the determining information has not been given, because no price compensates for underwriting blind.
One factor finally weighs which does not belong to the risk and which the insured can work on: the quality of the presentation. A file arriving complete, dated, with the ownership chain written out, the applicable treaties identified, the history of comparable measures documented and the difficult questions addressed in advance, is underwritten quickly and well. The same risk presented in three successive sendings, the third correcting the first, is underwritten badly or not at all, not out of formalism but because an underwriter obliged to assemble the file itself concludes that it does not know what it has not yet seen.
A remark on timing must be added because it costs to ignore. Underwriters do not have the same freedom all year: the first quarter follows treaty renewal, capacity is open and appetite wide; year end sees aggregation limits consumed and committees harder to persuade. Presenting a marginal file in November and re presenting it in February can be enough to change the answer, without any element of the risk having moved.
What this module teaches therefore comes down to a reversal. Stop presenting a file as a request for cover, and present it as an answer to the four questions the underwriter will ask itself anyway: what you carry elsewhere in this country, what I will be able to recover if this goes wrong, who you are when a file becomes difficult, and what you have not told me yet. A file built on those four questions does not guarantee acceptance; it saves both parties time and it removes the declines that were due only to presentation.
A broker presents on the same day, to the same underwriter, two files in the same Southeast Asian country, identically rated. The first concerns an electronic components plant held 100 per cent by a Japanese group through a Japanese subsidiary, limit sought 90 million euros, bilateral treaty in force between Japan and the country, a standard asset with a secondary market, a complete file in one sending including the ownership chain, the sector's regulatory history and unprompted mention of a tax audit closed in 2024. The second concerns a port concession held 60 per cent by a Luxembourg fund alongside a local partner, limit sought 85 million, no treaty between Luxembourg and the country, a unique asset with no comparable, a file arriving in three sendings the last of which corrects the capital split. The underwriter has already committed 140 million in that country against a 250 million aggregation limit. What will it do, and why?
It writes the first and declines the second, and almost nothing in that decision turns on the country, which is the same. The first sort is its portfolio: it carries 140 million against a 250 limit, so 110 remains and it cannot take both. That fact, invisible to both insureds, already eliminates one file before any examination, and it is what explains why a file declined here may be accepted elsewhere the same day. Then comes what the underwriter will be able to do if the event occurs, and the gap is clear. The Japanese file offers a bilateral treaty in force, a direct ownership chain allowing it to be invoked, and a standard asset whose value is established on a secondary market: on a loss, subrogation is worth something and quantum will not be argued for three years. The Luxembourg file offers no treaty, so no remedy of its own to transmit, and a unique asset whose value at the date of loss would be a dispute in itself. The first is rated as an advance of cash, the second as an outright loss, and the resulting premium gap exceeds what a buyer would accept. The third element bears on the insureds themselves and settles the decision. The Japanese group volunteered a closed tax audit, which costs it nothing and tells the underwriter how this file will be kept on the day it becomes difficult. The fund sent its file three times and the third version corrects the capital split, that is precisely the information determining whether a remedy exists: the underwriter will conclude, without saying so, that it does not know what it has not yet seen. What should be said to the fund is not that its country is bad, but that three things can be worked on before re presenting: routing the holding through an entity covered by a treaty, which is a structuring decision and not an insurance one; documenting the concession's value by a method the market recognises; and presenting in a single complete sending, in February rather than in November.
- 01The first sort is not about the file but about the underwriter's portfolio: that is why the same file receives three different answers from three markets on the same day.
- 02An underwriter does not look for a probability, it looks for what it will recover: treaty, ownership structure, recognisable value. It is subrogation reasoned before the loss.
- 03An unfavourable fact volunteered is worth more than a smooth file: it tells you about the person who will produce the documents on the day.
- 04What can be bounded is rated higher; what cannot be bounded, what no comparable allows to be modelled, and what is suspected of missing information, is declined.
- 05Timing counts without the risk having moved: capacity is open after treaty renewal and consumed at year end.