4 ms·

You analogy is incorrect as it would mean that she would buy the insurance when the flight is officially delayed. Here she is making a prediction, which may or

You analogy is incorrect as it would mean that she would buy the insurance when the flight is officially delayed. Here she is making a prediction, which may or may not be correct. It is exactly the same as when you are buying option or contract on the stock market. Here the issue is that the insurance company has a fixed premium which not based on the likelihood of delay. This is completely on the insurance side. Otherwise the stock market will have big issue if you apply the same analogy. This new clause is in effect a blanket one to not pay anything... if you pay for an insurance you have the conviction that the flight might be delayed... otherwise there would be no point to buy it....

roenxi 6y ago
The insurance company isn't a math problem; they are offering an explicitly bad deal. They are offering insurance on the basis that they will make a profit. Making that explicit in the contact is reasonable. If somebody thinks they are making money out of their insurance contract it is probably fraud or very close to.
- rswail 6y ago
  My understanding is that insurance companies operate on pricing the contracts to cover their potential losses while making profit on investing the insurance premiums.
  - Cogito 6y ago
    Insurance companies, depending on the company and the line of business, will aim to make both an underwriting profit and an investment profit. If it is a particularly high premium line, and especially if there is a lot of competition, it's more likely that the underwriting profit will be kept minimal or even negative, as the investment profit is so much more lucrative. This is common in the car insurance lines of business, for example. For retail insurers there are a lot of moving parts here, and they will often not hold much of the final risk, so can't make money from investment anyway. For example, the retailer (sometimes an underwriting agency) may be underwritten by some bigger insurer. That insurer may have a quota share arrangement with another insurer (I'll take 10% of the risk, you take 90%) with a corresponding premium share. At each level reinsurance is bought, sometimes on the whole portfolio of risks and sometimes on particular classes of risk or specifc policies, and the costs associated with the reinsurance are priced into the retail premium. Additionally, when working in an intermediated market with brokers, there are often credit terms of up to 90 days where no payments are made. Anecdotally, it does seem like some brokers do make a large amount of their money on this float. Adding it all together, there are two main numbers to track from an underwriting profit point of view - your loss ratio (claims incurred / earned premiums, often extended to include projections for future claims and premium for future insured periods) and your combined operating ratio (COR) which includes expenses. There will be a loss ratio at which a line of business becomes profitable, and final premiums are often calculated by taking the expected losses and grossing up by this loss ratio. For individual accounts it might literally be that simple - "over the last 5 years this account has averaged claims of $100,000 a year, we need a 60% loss ratio so the premium should be $166,667." If there are extra reinsurance costs you might add those, and you might also add a loading for large infrequent losses that aren't captured in your underlying loss ratio (and another loading for catastrophes, a level above that). The combined ratio needs to be under 100% to be 'underwriting profitable'. I've seen CORs as high as 105-110%, but typically those portfolios were being very closely monitored/remediated/exited. If there is a strong market the COR is allowed to drift higher, but when the market tightens so does the COR.
quietbritishjim 6y ago
Perhaps it's more likely getting life insurance when you know you have a disease that is likely to be fatal (but not 100% guaranteed). Most life insurance policies would disallow that. At the very least they require you to disclose it, and if they still agree to insure you the premium would be much higher.
unishark 6y ago
I'm guessing you don't have this woman's inside knowledge of the airline system in China. You're assuming here that a person can't reasonably know or deduce a flight will be delayed, but her track record says different. Sometimes, often in fact, it becomes pretty obvious a flight will be delayed but they just haven't bothered to officially announce the fact yet. Such as if the plane isn't even there because they are delayed on the previous leg, and boarding time is about to start. Or when the airport cancels all flights for an indeterminate time due to incoming bad weather and the airlines wait to update times because they don't know what to say yet.
- inimino 6y ago
  A certain flight out of a certain airport (I won't say which one) was routinely delayed to the point that passengers waiting for the inevitable compounding series of delay announcements (always after the predicted departure time had passed, and before the plane had showed up much less been boarded) would make small talk about it. Predicting that these flights would be delayed every. damn. time. was not difficult. However, if the insurance companies eliminate this face-saving measure for airlines then they may be forced to be realistic rather than maximally optimist in their departure estimates in the future (or simply be blacklisted by these kinds of insurance).
- greatpatton 6y ago
  Yes I do know by experience that it is very easy to deduce if a flight will be delayed. By experience in Europe, last leg of the day of an Easyjet flight -> 75% chance that your flight will be delayed. Snow during the night (look at the forecast) -> 90% your morning flight will be delayed etc. etc. No need for an advance ML model. Of course what she did was more like insider trading, but I'm pretty sure that you don't even need that.
  - user5994461 6y ago
    The difficulty to make money out of this might be to determine the delay with some accuracy, not just that there will be some delays. Flight are routinely delayed by tens of minutes (has a low-cost flight ever left on time?) and that won't award compensation. European laws define mandatory compensations when a flight is delayed or cancelled. This starts after 30 minutes if I remember well and depends how long the delay is and how long (distance) the flight is. Also, delay is measured on arrival time, not on departure time. Planes frequently depart late then arrive on time, because they fly faster to catch up.
notahacker 6y ago
There's a big difference between buying a flight ticket and insuring it because of the possibility of its delay increasing your costs which you want to avoid and buying a flight ticket and insuring it because of the probability of its delay earning you a reward which you are actively seeking out in your choice of flights It's the difference between buying flood disruption insurance because you operate in an occasionally flooded area, and moving your stock to wherever there's a flood warning. Particularly in this case, where the woman cancelled her expected on-time tickets because the tickets had no value to her except as a source of insurance payouts. Consumer insurance is not exactly the same as stock options: it's not supposed to be a perfectly efficient market. Insurers are often operating under rules which prevent them from denying or repricing insurance due to relevant risk factors and likewise premium holders are usually bound by far more terms and conditions preventing them from taking actions which might change the value of their policy than stocks or futures markets.

roenxi 6y ago

The insurance company isn't a math problem; they are offering an explicitly bad deal. They are offering insurance on the basis that they will make a profit. Making that explicit in the contact is reasonable. If somebody thinks they are making money out of their insurance contract it is probably fraud or very close to.

rswail 6y ago
My understanding is that insurance companies operate on pricing the contracts to cover their potential losses while making profit on investing the insurance premiums.
- Cogito 6y ago
  Insurance companies, depending on the company and the line of business, will aim to make both an underwriting profit and an investment profit. If it is a particularly high premium line, and especially if there is a lot of competition, it's more likely that the underwriting profit will be kept minimal or even negative, as the investment profit is so much more lucrative. This is common in the car insurance lines of business, for example. For retail insurers there are a lot of moving parts here, and they will often not hold much of the final risk, so can't make money from investment anyway. For example, the retailer (sometimes an underwriting agency) may be underwritten by some bigger insurer. That insurer may have a quota share arrangement with another insurer (I'll take 10% of the risk, you take 90%) with a corresponding premium share. At each level reinsurance is bought, sometimes on the whole portfolio of risks and sometimes on particular classes of risk or specifc policies, and the costs associated with the reinsurance are priced into the retail premium. Additionally, when working in an intermediated market with brokers, there are often credit terms of up to 90 days where no payments are made. Anecdotally, it does seem like some brokers do make a large amount of their money on this float. Adding it all together, there are two main numbers to track from an underwriting profit point of view - your loss ratio (claims incurred / earned premiums, often extended to include projections for future claims and premium for future insured periods) and your combined operating ratio (COR) which includes expenses. There will be a loss ratio at which a line of business becomes profitable, and final premiums are often calculated by taking the expected losses and grossing up by this loss ratio. For individual accounts it might literally be that simple - "over the last 5 years this account has averaged claims of $100,000 a year, we need a 60% loss ratio so the premium should be $166,667." If there are extra reinsurance costs you might add those, and you might also add a loading for large infrequent losses that aren't captured in your underlying loss ratio (and another loading for catastrophes, a level above that). The combined ratio needs to be under 100% to be 'underwriting profitable'. I've seen CORs as high as 105-110%, but typically those portfolios were being very closely monitored/remediated/exited. If there is a strong market the COR is allowed to drift higher, but when the market tightens so does the COR.

quietbritishjim 6y ago

Perhaps it's more likely getting life insurance when you know you have a disease that is likely to be fatal (but not 100% guaranteed). Most life insurance policies would disallow that. At the very least they require you to disclose it, and if they still agree to insure you the premium would be much higher.

unishark 6y ago

I'm guessing you don't have this woman's inside knowledge of the airline system in China. You're assuming here that a person can't reasonably know or deduce a flight will be delayed, but her track record says different. Sometimes, often in fact, it becomes pretty obvious a flight will be delayed but they just haven't bothered to officially announce the fact yet. Such as if the plane isn't even there because they are delayed on the previous leg, and boarding time is about to start. Or when the airport cancels all flights for an indeterminate time due to incoming bad weather and the airlines wait to update times because they don't know what to say yet.

inimino 6y ago
A certain flight out of a certain airport (I won't say which one) was routinely delayed to the point that passengers waiting for the inevitable compounding series of delay announcements (always after the predicted departure time had passed, and before the plane had showed up much less been boarded) would make small talk about it. Predicting that these flights would be delayed every. damn. time. was not difficult. However, if the insurance companies eliminate this face-saving measure for airlines then they may be forced to be realistic rather than maximally optimist in their departure estimates in the future (or simply be blacklisted by these kinds of insurance).
greatpatton 6y ago
Yes I do know by experience that it is very easy to deduce if a flight will be delayed. By experience in Europe, last leg of the day of an Easyjet flight -> 75% chance that your flight will be delayed. Snow during the night (look at the forecast) -> 90% your morning flight will be delayed etc. etc. No need for an advance ML model. Of course what she did was more like insider trading, but I'm pretty sure that you don't even need that.
- user5994461 6y ago
  The difficulty to make money out of this might be to determine the delay with some accuracy, not just that there will be some delays. Flight are routinely delayed by tens of minutes (has a low-cost flight ever left on time?) and that won't award compensation. European laws define mandatory compensations when a flight is delayed or cancelled. This starts after 30 minutes if I remember well and depends how long the delay is and how long (distance) the flight is. Also, delay is measured on arrival time, not on departure time. Planes frequently depart late then arrive on time, because they fly faster to catch up.

notahacker 6y ago

There's a big difference between buying a flight ticket and insuring it because of the possibility of its delay increasing your costs which you want to avoid and buying a flight ticket and insuring it because of the probability of its delay earning you a reward which you are actively seeking out in your choice of flights It's the difference between buying flood disruption insurance because you operate in an occasionally flooded area, and moving your stock to wherever there's a flood warning. Particularly in this case, where the woman cancelled her expected on-time tickets because the tickets had no value to her except as a source of insurance payouts. Consumer insurance is not exactly the same as stock options: it's not supposed to be a perfectly efficient market. Insurers are often operating under rules which prevent them from denying or repricing insurance due to relevant risk factors and likewise premium holders are usually bound by far more terms and conditions preventing them from taking actions which might change the value of their policy than stocks or futures markets.