BitShares Forum
Main => General Discussion => Topic started by: bytemaster on March 04, 2014, 07:31:26 am
-
There have been many discussions about Insurance, but today I spent time designing a workable DAC that will enable auto insurance, health insurance, life insurance, mudslide insurance, or even insurance against being caught committing a victimless crime. This system works by separating roles and responsibilities found in current insurance companies... current companies do the following tasks:
1) Price the risk (either over charging, or under covering) they never get it right.
2) Store the insurance funds ( they invest other peoples money, often poorly, to make up for miss pricing of risk)
3) They judge their own claims or make it hard to get payouts.
In other words, Insurance companies operate on fractional reserves just like they banks and make money selling insurance they are unable to make good on.
A DAC on the other hand must not make promises it cannot keep. Therefore, a DAC that implements an insurance system must only promise to pay claims to the extent supported by the market and no more.
Before signing up for the Insurance DAC it is important to understand exactly what you will receive for your money. What follows is a detailed explanation of how and when funds are distributed.
Defined Risk
Every insurance plan is based on the existance of some ratio of payout. If on any given month, 1 in 100 drivers may be in a car accident and the average accident cost $10,000 then the insurance premium would need to be $100 / month in order for the insurance company to break even.
Insurance companies are essentially gamblers who estimate the risk and make bets. They charge more for premiums than they estimate the risks warrant and keep the difference as their profit. Unfortunately, sometimes they gamble wrong and they can lose money. When you purchase a premium with an insurance company you have what is known as 'counter party risk' and in the event the insurance company bets wrong, you could be left holding the bag.
With the Insurance DAC, we do not attempt to estimate what the risk is nor do we make money by overcharging based on our estimate of the risk. We do not build up large pools of funds. Instead we operate on a Pay-as-You-Go basis.
Pay As You Go
In a perfectly balanced insurance system, the average total monthly payout should equal the average total monthly premiums paid by all customers. If this were not true the fund would either grow forever or become insolvant. Large pools of money are only necessary if claims are allowed to occur in lump sums that excede the monthly revenue. In the case of the Insurance DAC, large claims are paid out in installments up to the defined limit equal to total funds contributed divided by the risk. Risk can be calculated as the number of shares with a valid claim as a percentage of total shares times the average claim per share.
Each month a certain number of shares are owned by individuals whom have a qualifying claim. The revenue from contributions and collections for that month is divided evenly among the qualifying shares. Individuals who need more coverage can buy more shares.
If a particular individual has already received the maximum cumulative payout given their shares and the actual risks, then their 'shares' no longer qualify for additional monthly payments. If, over time, the average risk falls then they may resume receiving payments until their claim is paid off.
Only shares owned at the time the event (as determined by insurance adjuster) are considered when paying out a claim. If you purchase more shares after an incident, those shares only apply to future events. This is necessary to prevent people from loading up on shares after they are assured payout.
Insurance Adjustor
An insurance Adjuster is an individual who is granted authority by the market to award claims. They become licensed by the DAC by creating a BitAsset that tracks the total amount of outstanding awards they are allowed to grant. If the market trusts the adjustor, then they will increase their limits. Any awards made by the adjustor have a 30 day waiting period before they are paid out. If during the waiting period the market determines they have gone bad, then their limits will shrink and their most recent awards will be automatically canceled. An Insurance Adjustor makes money charging people fees (like a deductible) for hearing their claim. If the Adjustor considers the claim to be a fraud, they can simply reject it and keep the fees. The adjustor is also interested in minimizing the awards they issue so they can service as many people as possible (and thus collect more fees). If they payout a larger than necessary award they lose the opportunity to award other users. On the other hand, if they get a reputation for underpaying they will lose trust with the market participants who will simultainously switch to a more generous adjustor and reduce the total amount of awards that may be issued. In a sense, market forces will compel an Insurance Adjustor to be honest and fair in order to grow their business.
All Insurance premiums, awards, and payouts will be denominated in BitUSD or similar unit of account.
I recorded a 20 minute white board presentation on this idea which we will release soon. After significant review, we believe that this particular DAC has more potential than the other 5 DACs we have presented post-BitShares X. Just a heads up that we will be refining this DAC with a more detailed white paper and technical approach over the next year. It will have just as big an impact as BitShares X :)
-
Sounds fantastic! Bytemaster, can you be sure to check with legal counsel to make sure that no one involved would need state-level insurance licenses? I think it's on the same, slightly indirect 'hedging' level of Bitshares X. But I'd want to be double-sure it doesn't come under state regs.
I'm really excited about this DAC. Insurance companies make a killing by doing nothing more than what a pool of people can do by agreeing to combine their strength and cover each others' risks. This can work, help people save money, and be huge!
-
Sounds fantastic! Bytemaster, can you be sure to check with legal counsel to make sure that no one involved would need state-level insurance licenses? I think it's on the same, slightly indirect 'hedging' level of Bitshares X. But I'd want to be double-sure it doesn't come under state regs.
I'm really excited about this DAC. Insurance companies make a killing by doing nothing more than what a pool of people can do by agreeing to combine their strength and cover each others' risks. This can work, help people save money, and be huge!
I can assure you that we currently have multiple international law firms scrutinizing everything we plan to do. Over time our legal architecture is expected to evolve and adapt to survive and thrive in the dynamic regulatory environment.
-
Bytemaster, that is brilliant. I am so very glad to have found out about Invictus.
-
Awesome. Thanks for the assurance, Stan, and I appreciate your hard work. I, for one, am buying up PTS and AGS in anticipation of good things to come!
-
Ah shucks, I was hoping you would think of something clever for resolving claims. There's a whole class of DACs that you could do if you could get this right, but I'm reluctant to try until we get the low-hanging Entanglement-Free DACs first
-
Not sure if this is possible, but could the role of Adjustor be decentralized in some way too? Instead of a single person in the role of adjustor, could the approval of claims and the determination of the payout amount be crowd sourced?
For example, could multiple people purchase something like "Adjustor Shares" in an insurance DAC, which would grant them voting rights towards how each claim is handled? They could vote on things such as whether a claim is fraudulent, whether or not a claim should be approved, the amount of the award to be paid out, etc. Final payout amounts could be determined by averaging the amounts recommended by all adjustors, or something similar.
Participation could be encouraged by splitting fees only amongst the shareholders that actively voted on each claim. People that held "Adjustor Shares" but do not actively participate in voting over a length of time, could be "fined" a portion of their shares which could be awarded to the people that DO participate.
Fairness could also be encouraged by awarding Adjustors a percentage of the fees for each claim, based on how closely their votes were to the final average. People that cast far outlying votes (like voting for $0 payout on every claim) could also be "fined" a portion of their shares, which could be awarded to the "good" adjustors. Those fines should eventually weed out bad actors.
Could something like that work for an insurance-type DAC?
-
Sounds very cool!
-
While not the intention, this DAC can be extended to prediction/futures markets:
insurance that there won't be more than 5 named hurricanes in the Atlantic ocean this year
insurance that Brazil wins the world cup
insurance that "12 Years a Slave" wins the oscar for best picture
insurance that the DJIA rises >200 pts in April
etc.
-
Not sure if this is possible, but could the role of Adjustor be decentralized in some way too? Instead of a single person in the role of adjustor, could the approval of claims and the determination of the payout amount be crowd sourced?
For example, could multiple people purchase something like "Adjustor Shares" in an insurance DAC, which would grant them voting rights towards how each claim is handled? They could vote on things such as whether a claim is fraudulent, whether or not a claim should be approved, the amount of the award to be paid out, etc. Final payout amounts could be determined by averaging the amounts recommended by all adjustors, or something similar.
Participation could be encouraged by splitting fees only amongst the shareholders that actively voted on each claim. People that held "Adjustor Shares" but do not actively participate in voting over a length of time, could be "fined" a portion of their shares which could be awarded to the people that DO participate.
Fairness could also be encouraged by awarding Adjustors a percentage of the fees for each claim, based on how closely their votes were to the final average. People that cast far outlying votes (like voting for $0 payout on every claim) could also be "fined" a portion of their shares, which could be awarded to the "good" adjustors. Those fines should eventually weed out bad actors.
Could something like that work for an insurance-type DAC?
This...
Not sure if this is possible, but could the role of Adjustor be decentralized in some way too? Instead of a single person in the role of adjustor, could the approval of claims and the determination of the payout amount be crowd sourced?
For example, could multiple people purchase something like "Adjustor Shares" in an insurance DAC, which would grant them voting rights towards how each claim is handled? They could vote on things such as whether a claim is fraudulent, whether or not a claim should be approved, the amount of the award to be paid out, etc. Final payout amounts could be determined by averaging the amounts recommended by all adjustors, or something similar.
Participation could be encouraged by splitting fees only amongst the shareholders that actively voted on each claim. People that held "Adjustor Shares" but do not actively participate in voting over a length of time, could be "fined" a portion of their shares which could be awarded to the people that DO participate.
...would vastly decrease the efficiency of the system as a whole because all the adjusters have to validate the case as oposed to only one adjuster validating it. Or the quality of the validation is reduced when all the adjusters validate it only very superficially which makes the system vunerable to fraud.
While I think it is good to think about ways to decentralize things more it is important to see decentralization as a function (with the outcomes: trustlessness and efficiency) but not as a fashion.
A community vote could have a application for special claims which are highly subjective and are presented to the market/many adjusters by the full time adjuster if he thinks it needs comunity vote. Just an idea..
I like this one
Fairness could also be encouraged by awarding Adjustors a percentage of the fees for each claim, based on how closely their votes were to the final average.
...maybe with two adjusters and for critical cases?
-
There have been many discussions about Insurance, but today I spent time designing a workable DAC that will enable auto insurance, health insurance, life insurance, mudslide insurance, or even insurance against being caught committing a victimless crime. This system works by separating roles and responsibilities found in current insurance companies... current companies do the following tasks:
[....]
I recorded a 20 minute white board presentation on this idea which we will release soon. After significant review, we believe that this particular DAC has more potential than the other 5 DACs we have presented post-BitShares X. Just a heads up that we will be refining this DAC with a more detailed white paper and technical approach over the next year. It will have just as big an impact as BitShares X :)
Who is (are) the market (participants) mentioned in the section about the Insurance adjuster? Is the market here all kinds of people or is it the share holders?
And would the share holders be paid dividens?
-
There have been many discussions about Insurance, but today I spent time designing a workable DAC that will enable auto insurance, health insurance, life insurance, mudslide insurance, or even insurance against being caught committing a victimless crime. This system works by separating roles and responsibilities found in current insurance companies... current companies do the following tasks:
[....]
I recorded a 20 minute white board presentation on this idea which we will release soon. After significant review, we believe that this particular DAC has more potential than the other 5 DACs we have presented post-BitShares X. Just a heads up that we will be refining this DAC with a more detailed white paper and technical approach over the next year. It will have just as big an impact as BitShares X :)
Who is (are) the market (participants) mentioned in the section about the Insurance adjuster? Is the market here all kinds of people or is it the share holders?
And would the share holders be paid dividends?
Most DACs have three overlapping markets:
Those who want its currency so they can purchase its services.
Those who want its currency because they get a share of its profits.
Those who want its currency because they think it will appreciate.
Demand for the currency for any combination of these reasons drives its price.
Holders of the currency for any combination of these reasons can get all three benefits.
Built-in bitAssets (e.g bitUSD) allow a holder to select which benefit(s) they claim.
-
There have been many discussions about Insurance, but today I spent time designing a workable DAC that will enable auto insurance, health insurance, life insurance, mudslide insurance, or even insurance against being caught committing a victimless crime. This system works by separating roles and responsibilities found in current insurance companies... current companies do the following tasks:
[....]
I recorded a 20 minute white board presentation on this idea which we will release soon. After significant review, we believe that this particular DAC has more potential than the other 5 DACs we have presented post-BitShares X. Just a heads up that we will be refining this DAC with a more detailed white paper and technical approach over the next year. It will have just as big an impact as BitShares X :)
Who is (are) the market (participants) mentioned in the section about the Insurance adjuster? Is the market here all kinds of people or is it the share holders?
And would the share holders be paid dividends?
Most DACs have three overlapping markets:
Those who want its currency so they can purchase its services.
Those who want its currency because they get a share of its profits.
Those who want its currency because they think it will appreciate.
Demand for the currency for any combination of these reasons drives its price.
Holders of the currency for any combination of these reasons can get all three benefits.
Built-in bitAssets (e.g bitUSD) allow a holder to select which benefit(s) they claim.
Thanks Stan.
What I meant specifically was, who are the market participants that "judge" / "evaluate" the adjuster's performance/credibility? Anybody or the insurance share holders?
-
There have been many discussions about Insurance, but today I spent time designing a workable DAC that will enable auto insurance, health insurance, life insurance, mudslide insurance, or even insurance against being caught committing a victimless crime. This system works by separating roles and responsibilities found in current insurance companies... current companies do the following tasks:
[....]
I recorded a 20 minute white board presentation on this idea which we will release soon. After significant review, we believe that this particular DAC has more potential than the other 5 DACs we have presented post-BitShares X. Just a heads up that we will be refining this DAC with a more detailed white paper and technical approach over the next year. It will have just as big an impact as BitShares X :)
Who is (are) the market (participants) mentioned in the section about the Insurance adjuster? Is the market here all kinds of people or is it the share holders?
And would the share holders be paid dividends?
Most DACs have three overlapping markets:
Those who want its currency so they can purchase its services.
Those who want its currency because they get a share of its profits.
Those who want its currency because they think it will appreciate.
Demand for the currency for any combination of these reasons drives its price.
Holders of the currency for any combination of these reasons can get all three benefits.
Built-in bitAssets (e.g bitUSD) allow a holder to select which benefit(s) they claim.
Thanks Stan.
What I meant specifically was, who are the market participants that "judge" / "evaluate" the adjuster's performance/credibility? Anybody or the insurance share holders?
Anyone is eligible since the system is open to all. In practice, its only people who think they can make money by specializing in assessing and speculating on the reputations of others.
The number of new careers that can be followed from a beach in Bora Bora is mind boggling.
(http://0.tqn.com/d/create/1/5/w/k/8/-/Le-Meridien-Bora-Bora---Aerial.jpg)
-
That's right. These DACs may cost a lot of jobs in the short term, all of them in the greedy and wasteful financial sector. But oh, the opportunities it will create! If we think eBay and the Amazon storefront created a lot of new chances for do-it-yourself businesses, we ain't seen nothing' yet! There will be many, many new opportunities for people to make great incomes...from anywhere in the world that's connected.
-
If someone does go on holiday there, to the place above, and they profitably traded Bitshares whilst there to pay for the trip, it would be epic advertising. Obviously it would need to be an independent forum member, or possibly someone with no trading skills to emphasize ease of use.
Sent from my iPad using Tapatalk (http://tapatalk.com/m?id=1)
-
That's right. These DACs may cost a lot of jobs in the short term, all of them in the greedy and wasteful financial sector. But oh, the opportunities it will create! If we think eBay and the Amazon storefront created a lot of new chances for do-it-yourself businesses, we ain't seen nothing' yet! There will be many, many new opportunities for people to make great incomes...from anywhere in the world that's connected.
Agree, there will be a lot of new oportunities and traditional careers are 20th century. But overall the gloabl tendency towards automatization combined with economies of scale (which is interconnected) will decrease the need for employees/working staff drastically. There will be very well paid individuals (those allocating capital and those maintaining systems) and a lot of unskilled unemployed and low skilled low paid people. The reason is simple. If the overall need for employees decreases there are more that can (and have to if they want to survive) do the jobs which dont need high qualification. The need for efficiency and the fact of competition in a free market plus the demand for profit will make employers pay the least possible price/loan for those jobs for which work force can easily be replaced.
So not everything is rosy for everyone but all this is part of a tendency ongoing since mankind began to cultivate land. Division of labour, individualization and automatization are all part of the same trend.
It is interesting to ask what this crypto/blockchain technology actually decentralizes in the above context (the one of the last sentence). It decentralizes a task to many suppliers (blockchains are secured through the efforts of many, ideally ;-) ; "adjuster speculators" are many and access the adjuster together) but blockchain technology and prediction markes centralize those tasks insofar as the individuals/machines that do theses tasks mostly do nothing but those tasks and have to specialize to survive in the competetive market they find themselfs in. Insofar decentralization is also part of that global trend. [This is a tendency; there surely are counter examples]
This trend can not be avoided given the prevailing conditions (free markets, population big/dens enough to allow it and require it, rational/material mindset fostering it (might not be a necessary condition)). But it will defenitely confront humanity with unseen challenges that are best discussed with an open mind.
-
I see no way this could possibly work.
How can the market get enough information about an adjuster to evaluate whether he/she is fraudulent?
-
I see no way this could possibly work.
How can the market get enough information about an adjuster to evaluate whether he/she is fraudulent?
It could possibly work for life insurance where the outcome is pretty well beyond dispute. It might also be feasible in a small enough group where all the members are known to each other (say a church group).
-
The number of new careers that can be followed from a beach in Bora Bora is mind boggling.
(http://0.tqn.com/d/create/1/5/w/k/8/-/Le-Meridien-Bora-Bora---Aerial.jpg)
Damn you, now I have an overwhelming desire to escape to bora bora
-
I see no way this could possibly work.
How can the market get enough information about an adjuster to evaluate whether he/she is fraudulent?
Sure it could.
Let's take car insurance. Comprehensive car insurance.
You rear end a BMW. You call the police. You get the police report. You send proof of car ownership, police report, multiple estimates to fix both cars and proof of insurance before and after date of accident to 3 different paid "adjusters" or "arbitrators" (who never have to leave the computer) and as long as 2 agree on one way or the other, that decision is final.
For scratches or car hit by shopping cart you would need to submit verified/dated photos or videos of car before coverage began and then after damage plus estimates to fix.
Health insurance might be tougher since a Dr. Might not see you because he doesn't accept BTS insurance.
But you could do E.R. Insurance. Buy the insurance before emergency room visit. Send bill to multiple claims adjusters to analyze and make decision.
The only thing to get around is that you need to be able to force people to renew. You can't just have someone pay $100 worth of BTS insurance and be fully covered for life. So there might need to be some kind of burn or expiration to every share.
Does any of this make sense? Am I seeing this right?
-
Another thought:
Is there a way to have both an "insurance share" and an "investor share"?
Say 100 people want to buy 1 "investor share" each at $100 (using dollars because it's easier to visualize). You now added a $10,000 reserve to the insurance pot that also has all the monthly premiums. If total premium collected is $100,000, you now have $110,000 in the pool. If the claims for that month are $90,000 then $10000 is paid as a dividend back to the investors who now have a chance of earning money next month. Now if the next month you have the same $100,000 in premiums and $10,000 in investment reserves and the claims exceed $100k the investor gets no dividend. But if the following month $100k premium and $50k in claims, then $50k is split as a dividend to investors in the form of $5k each.
Of course these numbers could all be different and dividend payout could be a certain percentage of the unclaimed premiums, say 25%. That way it would add more to the following months reserve and so on....
You then could sell more shares each month that receive less dividend than the previous month so as to not dilute the original investors. Or only sell a limited number of shares per premium collected each month. And if premium drops, payouts will drop and investors could sell their share on a secondary market.
Is this possible?
-
That is an interesting concept.
Sent from my iPhone using Tapatalk (http://tapatalk.com/m?id=1)
-
Basically, the insured is betting that he or she will suffer some cost. Insurance works, because the cost is something the people will tend to avoid, like broken bones, wrecked cars, and such.
In the scenario above, the investors are betting that the insured are excessively pessimistic.
-
Basically, the insured is betting that he or she will suffer some cost. Insurance works, because the cost is something the people will tend to avoid, like broken bones, wrecked cars, and such.
In the scenario above, the investors are betting that the insured are excessively pessimistic.
Isn't that what insurance companies do? They hope to collect more in premiums than they payout in claims. If you end up with reserves and the premium to payout ratio is more than 1, the "investors" or owners of the insurance company make a profit.
Agencies do something similar where they pool their premiums together and those who have less than a certain percentage of the premiums in claims get a profit and those who exceed the threshold get nothing. They are spreading the risk amongst more agencies for a better chance at steady profit.
-
Basically, the insured is betting that he or she will suffer some cost. Insurance works, because the cost is something the people will tend to avoid, like broken bones, wrecked cars, and such.
In the scenario above, the investors are betting that the insured are excessively pessimistic.
Isn't that what insurance companies do? They hope to collect more in premiums than they payout in claims. If you end up with reserves and the premium to payout ratio is more than 1, the "investors" or owners of the insurance company make a profit.
Agencies do something similar where they pool their premiums together and those who have less than a certain percentage of the premiums in claims get a profit and those who exceed the threshold get nothing. They are spreading the risk amongst more agencies for a better chance at steady profit.
If the insurer has fixed costs of anything greater than zero, then it must collect more in premiums than it pays out in claims.
The goal here is not so much to compete head-to-head with AIG, Allianz, or Allstate within OECD member states, as to bring risk mitigation to some of the six billion who live Developing World countries, and to those within OECD members states who do not have access to certain kinds of insurance.
My personal interest in this involves vendor credit guarantees.
-
+5% agreed, the waters in this DAC are best tested on the most predicable risk. built in well established methods to assess risk are essential to attract mainstream investor and promote adoption.
-
Has a developer agreed to take this on yet? Is it something Invictus wants to develop or is it going to be a third party thing?
This is the single DAC I'm most excited about.
-
http://bitcoinist.net/umbrella-holdings-insure-your-cryptocurrencies/ it seems it's the first insurance related crypto? Although centralized, just wanted to share it here, it might spark some ideas, who knows.