I'm in India, and I'm currently trying to understand the accounting (i.e. simple arithmetic) of the Public Disclosure information of Indian insurance companies. The context is that I'm trying to figure out what health insurance policy I should buy. It's also useful to be able to read and understand such statements.

I realise that most people on the site are not Indian, but I hope that the way these companies present their information is sufficiently similar to other places that non-Indians can make sense of it. Because I am currently having difficulty doing so.

Here are two documents which I downloaded off the web. The first is:


The second is:


I'm interested in health insurance, so consider the following data about IFFCO Tokio from both these sources. NOTE: I'm not specially interested in IFFCO Tokio; I'm just using it as an example.

The first, "GENERAL INSURANCE CLAIM INSIGHTS A POLICYHOLDERS HANDBOOK - 4th Edition" is a handbook produced by an organization calledInsurance Brokers Assocation of India (IBAI), and appears to be designed to help people make a decision about which insurer to buy found this PDF online. It does not appear to be on the organization's web site, though earlier versions of this handbook are. The handbook does not cite its sources for the published statistics. It just mentions the term "based on public disclosure by each insurer themselves on their website as required under IRDAI's public disclosure of data requirements." See Methodology (page 5 of the handbook).

Some statistics for IFFCO Tokio from this handbook are as follows.(Percentages only - no counts are given.)

(From pg no 65 of the handbook)

96.57% (for year ended March 31, 2019)     3,37,020

(pg no 69)

24.64% (for year ended March 31, 2019)

(pg no 71)

3.77% (for year ended March 31, 2019)

The percentages make little sense to me. From the definitions, it looks like it should be true that:


but that is not the case. Not even close.

96.57% + 24.64% + 3.77% = 124.98% 

So what am I not understanding about these definitions?

The second example is from

https://www.iffcotokio.co.in/content/dam/iffcotokio/iffco-pdf/sites/default/files/Public%20Disclosure%20March%202019_0.pdf which is a public disclosure document from IFFCO Tokio for the 2018-2019 financial year.

From Page 47 of 72

FORM NL-25 : Quarterly claims data for Non-Life (Column 9) Date: 1st Apr'18 to 31st Mar'19

(just the Health column)

Claims O/S at the beginning of the period    27605 
Claims reported during the period           309415 
Claims Settled during the period            325468
Claims Repudiated during the period          33166
Claims closed during the period                  0 
Claims O/S at End of the period              12707 
Less than 3 months                            7522 
3 months to 6 months                          2754 
6 months to 1 year                            2187
1 year and above                               244

Comments: I assume O/S means outstanding, though I don't know why they can't just write outstanding.

Based on what I know, it seems reasonable to assume that: Claims O/S at the beginning of the period + Claims reported during the Claims Settled during the period + Claims Repudiated during the period + Claims closed during the period + Claims O/S at End of the period

LHS = 27605 + 309415 = 337020
RHS = 325468 + 33166 + 0 + 12707 = 371341

These two aren't equal, though.

371341.0/337020 =  1.1018366862500741

What am I missing?

On the plus side, at least

7522 + 2754 + 2187 + 244 = 12707.

1 Answer 1


I am not entirely sure how to understand this although I worked some time in the industry in Germany.

There are two ways to go about it:

  1. Repudiated claims are not considered valid claims, and have to be subtracted from the total number of claims, before calculating the rest.
  2. Repudiation is a form of settlement, so the Repudiated claims are contained in the number of settled claims

The document actually gives you the formula and the total value of cases (337,020) so I think they have gone with option 2.

of the 337,020 claims, 24.64% are repudiated = 83,042
of the 337,020 claims, 96.57% are settled = 325,460
So claims actually paid to are 325,460-83,042 = 242.418‬ 

Either way settled: 96.57% + outstanding: 3.77% = 100.37% (should be 100% from my viewpoint you can see where this comes from in the iffco-report).

So the % settled gives you an idea of how fast a insurer will react to your claims, the repudiation ratio gives you an idea of how likely that reaction will result in a payout.

The IFFCO report:

Health Claims O/S at the beginning of the period 27,605 
Claims reported during the period 309,415 
Claims Settled during the period 325,468 
Claims Repudiated during the period 33,166 
Claims closed during the period 0 
Claims O/S at End of the period 12,707


Total claims in period 27,605 + 309,415 = 337.020‬
- Settled 325,468 claims (33,166 of which are repudiated)  = 11.552 

So I think the Claims O/S at End of the period is slightly off, which explains the slightly off percentage in the other report.

Note that both reports give us 337,020 as total number of claims. Also the settlement ratio is the same 96,57% The slightly more then 100% settled + outstanding can be explained by the oddly inflated number of outstanding claims. I can´t seem to make sense of the 24% repudiated claims vs the number given by IFFCO, 33,166.

  • Hi @Daniel. Thank you for your answer. It's very helpful. So, if I understand correctly, the actual number for repudiated claims is in fact 33166.0/337020 = 0.0984. In other words, 9.84%, not 24.64%? Also 325468 + 12707, which is not the same as 337,020. The difference is 1155. Apr 3, 2020 at 13:00
  • So it seems that I can't trust this IBAI compilation of data. It's nice that at least some of the data matches, otherwise that would be really worrisome. It's also odd that IFFCO Tokio's own data does not appear to be internally consistent, unless I'm misunderstanding something. Also, that's not the meaning of "Settled" that is normally given in discussions on the subject. Usually, settled is said to mean that the claim was paid out. Apr 3, 2020 at 13:02
  • Good question regarding the first document would be how they came to the 24.64% repudiation rate. This is a severe difference and one that could potentially impact negatively on future sales of IFFCO if that is some sort of widely used document. So I´d recon they´d want to get this corrected? Or is there something we both don´t understand?
    – Daniel
    Apr 3, 2020 at 15:24
  • Regarding the internal consistency of the IFFCO-Data ... from what I´ve seen so far i my career Insurance data is often messy, claims especially so but I would find it still a little bit odd if they would not even care to clean this up for a official publication. Might be worthwhile to ask them directly...
    – Daniel
    Apr 3, 2020 at 15:27
  • Hi Daniel, please excuse the late reply. I doubt many people look at that IBAI handbook. The only information on IBAI's site is the accouncement of an "event" during which they released the handbook. The handbook itself does not seem to be there, as I already said. Nor is any information about purchasing it. Maybe it was put together by interns? It's lacking a lot of basic information that would make it really useful. I actually already wrote to IBAI, but did not get a reply. In any case, I could write to them again pointing out the error. Apr 5, 2020 at 10:21

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .