Risk Management 2.0
Matthew Maddocks discusses one possible view regarding the future of risk management for specialty (re)insurers , and why not to follow the property catastrophe path
Specialty (re)insurers face a number of different challenges to property catastrophe (re)insurers. In terms of sophistication and analysis capability they are considered the poor relation, but they have an opportunity now to do things differently, to embrace modern working practices and technology and make risk management a competitive advantage rather than purely a reporting function.
Risk Management 101
The most basic definition of risk management, for a (re)insurance company, is the ability to know the financial impact on the company for a given loss event before it happens. With this information management decisions can be made to ensure that the company will continue trading if this loss event occurs. This is obviously one of the most fundamental and basic functions performed by a company but with the added bonus that, if done correctly, gives the ability to price the individual risks and portfolio being assumed.
For (re)insurers there are therefore 2 distinct requirements for a complete risk management process;
(1) the ability to model the types of loss events that are likely to occur in the future (including the expected frequency), and
(2) the ability to know the loss that a company will incur as a result of or event happening.
We’ll return to these functions below but it’s worth noting here that only (1) above is subjective and that in a perfect world almost no competitive advantage can be gained by performing (2) accurately.
The Past and (Sometimes) Present
For specialty (re)insurers, in the past, risk management would either have been based around overestimation of exposures with no PML (probable maximum loss) consideration, or around outdated client exposures stored and analysed in spreadsheets that gave “roughly” the correct answer. Very little effort was put into modeling losses (as in (1) above) with most taking a realistic disaster scenario approach and looking at only a handful of possible events, with no attempt to estimate the likely frequency of the event happening. This approach might have been conservative, but it is not particularly informative and would not help a company in efficiently utilising its capital base.
For (re)insurers, the traditional approach was as a result of; a lack of transparency in the insurer-reinsurer relationship, lack of standardised data formats, inefficient data transmission (i.e. email from insurer to broker, email from broker to reinsurer), and a lack of a software solution to store, process, and analyse the data. As we’ll see I think these impediments are slowing disappearing, and in the future will no longer prevent risk managers from doing their job as effectively as they otherwise could.
The Future - Risk Management 2.0
The future of risk management, I think, will be governed by two overarching principles: firstly market collaboration where collaboration makes sense; and secondly (re)insurers investing in their own competitive advantage.
Data
Increasingly, across all industries, data is becoming easier to access and cheaper to obtain. With advanced web technologies to facilitate crowd-sourced data, traditional notions of proprietary data are disappearing. If you’re not convinced ask the CEO at Encyclopedia Britannia how their business has been impacted by the rise of Wikipedia. In the future information barriers will be dramatically reduced, as ‘free-market’ data is created, something that risk managers should not only be aware of but should actively embrace.
In practice what does this mean? The best way is probably to give an example from one of the specialty classes, say satellite (re)insurance. If all insurers and reinsurers agreed to create a wiki-style database detailing satellite characteristics (make, model, launch details, etc) then the market would have an incredibly powerful and free resource. In addition to the cost savings that could be achieved, the great thing about crowd-sourced data/information is that it likely to be highly accurate due to market peer review. This common approach would also, by definition, create a data template for storing and transmitting satellite data, removing many of the issues surrounding mapping such as naming conventions and field formats.
The second point that needs to be raised is the data regarding insurers’ exposure: a vital element in reinsurer risk management. Many insurers are concerned with sharing their questionnaire (a table detailing the insurance policies their reinsurance programmes are protecting) but this again is something that I think will change. Insurers that are prepared to be more transparent with their information will undoubtedly succeed where others will fail. If I were a satellite insurer I would be happy for all genuine reinsurers in this class to see my exposures, confident that I would end up with a more efficient reinsurance programme without in any way jeopardising my competitive advantage.[2]
This “free market” approach to data is in stark contrast to that found in the property catastrophe market where closed data standards, incompatible formats, and “black-box” data and analysis are commonplace. None of these are particularly desirable from a (re)insurer point of view and only lead to increased cost, administrative problems, and business models that are based on an inefficient market - something that is ultimately unsustainable in the long-run.
Loss Modelling
Loss modeling (as in (1) above) is probably the single largest competitive advantage that a company can have over its competitors. Understanding which risks are more or less likely to have a loss, and the expected distribution of the severity of the loss, will ultimately mean your return on equity will be higher or lower than the next company. This is the area that risk managers should be investing in, an investment that will potentially see a higher rate of return than any other.
In the future as risk trading becomes more instantaneous there will be increasing opportunities for companies to benefit from understanding losses (and therefore risk) better than their competitors. Like most other financial markets, mechanisms will be developed for companies to trade on these different views on a daily basis, something that will see the better companies rise to the top of the market while removing those companies that do not know how to correctly price or manage risk.
Again this is not the approach that the property catastrophe market has taken. Most (re)insurance companies use vendor models to analyse market losses, with the software telling them the relative frequency of losses, the risks that will impacted, and the severity of the loss. Although some parameterisation is possible the models fundamentally provide the market with a standardised approach. I think the future of specialty risk management will be one where companies demand technology solutions that allow them to fully model their view of the world, with those that invest in this approach gaining a real competitive advantage over their peers.
Solvency II
Solvency II should impact both the data and loss modeling issues discussed above. The new regime should encourage a greater transparency between insurers and reinsurers, and should also focus risk managers attention on loss modeling at a line of business level and aggregating them at the company level. The 101 definition of risk management above (which admittedly is only concerned with underwriting risk) would be at the heart of any enterprise-wide capital model. Unless this basic definition is satisfied any additional complexity in the model would be relatively useless.
Risk Management Clouds
The final thing for discussion is the future of technology in the risk management sector. Cloud, or hosted, solutions will undoubtedly become more commonplace as third party companies provide secure and extremely powerful processing technology, at a much more competitive price than could be achieved internally. This, combined with tight legal agreements should mean that concerns regarding confidentiality are a thing of the past. In addition to base storage and processing, these risk management clouds could also seamlessly integrate with crowd-sourced data and insurer questionnaire repositories, the result being an incredibly automated and efficient risk management process.
Conclusion
The future of risk management should be clear. Practitioners should embrace collaborative ways of workings, while at the same time focusing internally on that which will add true value to their company and enable them to outperform their competitors. The bar will undoubtedly be raised for the entire industry and some companies will be able to “free-ride” on these developments, but ultimately a free-rider will never threaten those companies that are true market leaders.
Matthew Maddocks is Product Manager for Russell Group
Specialty (re)insurers face a number of different challenges to property catastrophe (re)insurers. In terms of sophistication and analysis capability they are considered the poor relation, but they have an opportunity now to do things differently, to embrace modern working practices and technology and make risk management a competitive advantage rather than purely a reporting function.
Risk Management 101
The most basic definition of risk management, for a (re)insurance company, is the ability to know the financial impact on the company for a given loss event before it happens. With this information management decisions can be made to ensure that the company will continue trading if this loss event occurs. This is obviously one of the most fundamental and basic functions performed by a company but with the added bonus that, if done correctly, gives the ability to price the individual risks and portfolio being assumed.
For (re)insurers there are therefore 2 distinct requirements for a complete risk management process;
(1) the ability to model the types of loss events that are likely to occur in the future (including the expected frequency), and
(2) the ability to know the loss that a company will incur as a result of or event happening.
We’ll return to these functions below but it’s worth noting here that only (1) above is subjective and that in a perfect world almost no competitive advantage can be gained by performing (2) accurately.
The Past and (Sometimes) Present
For specialty (re)insurers, in the past, risk management would either have been based around overestimation of exposures with no PML (probable maximum loss) consideration, or around outdated client exposures stored and analysed in spreadsheets that gave “roughly” the correct answer. Very little effort was put into modeling losses (as in (1) above) with most taking a realistic disaster scenario approach and looking at only a handful of possible events, with no attempt to estimate the likely frequency of the event happening. This approach might have been conservative, but it is not particularly informative and would not help a company in efficiently utilising its capital base.
For (re)insurers, the traditional approach was as a result of; a lack of transparency in the insurer-reinsurer relationship, lack of standardised data formats, inefficient data transmission (i.e. email from insurer to broker, email from broker to reinsurer), and a lack of a software solution to store, process, and analyse the data. As we’ll see I think these impediments are slowing disappearing, and in the future will no longer prevent risk managers from doing their job as effectively as they otherwise could.
The Future - Risk Management 2.0
The future of risk management, I think, will be governed by two overarching principles: firstly market collaboration where collaboration makes sense; and secondly (re)insurers investing in their own competitive advantage.
Data
Increasingly, across all industries, data is becoming easier to access and cheaper to obtain. With advanced web technologies to facilitate crowd-sourced data, traditional notions of proprietary data are disappearing. If you’re not convinced ask the CEO at Encyclopedia Britannia how their business has been impacted by the rise of Wikipedia. In the future information barriers will be dramatically reduced, as ‘free-market’ data is created, something that risk managers should not only be aware of but should actively embrace.
In practice what does this mean? The best way is probably to give an example from one of the specialty classes, say satellite (re)insurance. If all insurers and reinsurers agreed to create a wiki-style database detailing satellite characteristics (make, model, launch details, etc) then the market would have an incredibly powerful and free resource. In addition to the cost savings that could be achieved, the great thing about crowd-sourced data/information is that it likely to be highly accurate due to market peer review. This common approach would also, by definition, create a data template for storing and transmitting satellite data, removing many of the issues surrounding mapping such as naming conventions and field formats.
The second point that needs to be raised is the data regarding insurers’ exposure: a vital element in reinsurer risk management. Many insurers are concerned with sharing their questionnaire (a table detailing the insurance policies their reinsurance programmes are protecting) but this again is something that I think will change. Insurers that are prepared to be more transparent with their information will undoubtedly succeed where others will fail. If I were a satellite insurer I would be happy for all genuine reinsurers in this class to see my exposures, confident that I would end up with a more efficient reinsurance programme without in any way jeopardising my competitive advantage.[2]
This “free market” approach to data is in stark contrast to that found in the property catastrophe market where closed data standards, incompatible formats, and “black-box” data and analysis are commonplace. None of these are particularly desirable from a (re)insurer point of view and only lead to increased cost, administrative problems, and business models that are based on an inefficient market - something that is ultimately unsustainable in the long-run.
Loss Modelling
Loss modeling (as in (1) above) is probably the single largest competitive advantage that a company can have over its competitors. Understanding which risks are more or less likely to have a loss, and the expected distribution of the severity of the loss, will ultimately mean your return on equity will be higher or lower than the next company. This is the area that risk managers should be investing in, an investment that will potentially see a higher rate of return than any other.
In the future as risk trading becomes more instantaneous there will be increasing opportunities for companies to benefit from understanding losses (and therefore risk) better than their competitors. Like most other financial markets, mechanisms will be developed for companies to trade on these different views on a daily basis, something that will see the better companies rise to the top of the market while removing those companies that do not know how to correctly price or manage risk.
Again this is not the approach that the property catastrophe market has taken. Most (re)insurance companies use vendor models to analyse market losses, with the software telling them the relative frequency of losses, the risks that will impacted, and the severity of the loss. Although some parameterisation is possible the models fundamentally provide the market with a standardised approach. I think the future of specialty risk management will be one where companies demand technology solutions that allow them to fully model their view of the world, with those that invest in this approach gaining a real competitive advantage over their peers.
Solvency II
Solvency II should impact both the data and loss modeling issues discussed above. The new regime should encourage a greater transparency between insurers and reinsurers, and should also focus risk managers attention on loss modeling at a line of business level and aggregating them at the company level. The 101 definition of risk management above (which admittedly is only concerned with underwriting risk) would be at the heart of any enterprise-wide capital model. Unless this basic definition is satisfied any additional complexity in the model would be relatively useless.
Risk Management Clouds
The final thing for discussion is the future of technology in the risk management sector. Cloud, or hosted, solutions will undoubtedly become more commonplace as third party companies provide secure and extremely powerful processing technology, at a much more competitive price than could be achieved internally. This, combined with tight legal agreements should mean that concerns regarding confidentiality are a thing of the past. In addition to base storage and processing, these risk management clouds could also seamlessly integrate with crowd-sourced data and insurer questionnaire repositories, the result being an incredibly automated and efficient risk management process.
Conclusion
The future of risk management should be clear. Practitioners should embrace collaborative ways of workings, while at the same time focusing internally on that which will add true value to their company and enable them to outperform their competitors. The bar will undoubtedly be raised for the entire industry and some companies will be able to “free-ride” on these developments, but ultimately a free-rider will never threaten those companies that are true market leaders.
Matthew Maddocks is Product Manager for Russell Group
Labels: Product Announcement, Thought Leadership
posted by Suki Basi at
00:54
0 Comments:
Post a Comment
Subscribe to Post Comments [Atom]