February 8, 2018
VIA EDGAR
United States Securities and Exchange Commission
Division of Corporate Finance
Office of Healthcare & Insurance
Washington, D.C. 20549-3628
Attention: Rolf Sundwall / Kevin Vaughn
Re: | Greenlight Capital Re, Ltd. |
Form 10-K for the Fiscal Year Ended December 31, 2016 (the “Form 10-K”)
Form 10-Q for the Quarterly Period Ended September 30, 2017 (the “Form 10-Q”)
File No. 001-33493
Dear Sirs:
Set forth below is the response of Greenlight Capital Re, Ltd. (the "Company", "our" or "we") to the comment letter of the staff (the "Staff") of the Securities and Exchange Commission (the "Commission") dated January 25, 2018 with respect to the Form 10-K and Form 10-Q referenced above.
For your convenience, we have set forth below the Staff's comments contained in your letter referenced above followed by the Company's response thereto. Caption references and page numbers refer to the captions and pages contained in the Form 10-K or Form 10-Q, as applicable, unless otherwise indicated. Capitalized terms used but not otherwise defined herein have the meanings ascribed to such terms in the Form 10-K or Form 10-Q, as applicable. All numbers are in thousands, except where noted otherwise.
Form 10-K for Fiscal Year Ended December 31, 2016
Notes to the Consolidated Financial Statements
7. Loss and Loss Adjustment Expense Reserves
Disclosures about Short Duration Contracts, page F-32
1. | Refer to your response to our prior comment 3. In order for us to more fully evaluate the appropriateness of your aggregation of the lines included in Appendix A into the frequency, severity and health categories, please tell us which of the lines or parts thereof are included in each category. Further, provide us an analysis, for each category, of the characteristics of the lines and parts thereof demonstrating that they do not have significantly different characteristics. Refer to ASC 944-40-50-4H. |
As noted previously, we allocate our business into two categories: frequency and severity. We do not allocate business to frequency or severity based upon line of business. We consider each deal individually as part of our pricing process and allocate deals into frequency or severity based upon several factors. These factors include, but are not limited to:
1
– | Policy limits of the underlying business; |
– | Attachment points of the underlying policies; |
– | Historical volatility and its causes within the loss experience |
As far as whether the characteristics of each deal or line of business are sufficiently similar, it is important first to clarify what characteristics are being discussed/considered. Insurance and reinsurance exposures have, of course, many characteristics that could be of interest depending on the intended use of the reader. In our experience, these characteristics include:
– | The volume and frequency of claims occurring |
– | The likelihood and prevalence of large claims |
– | Claim reporting delays |
– | Claim settlement delays |
– | The volatility of initial loss estimates on reported claims relative to the final claim settlement figure (particularly where the final outcome could result in no loss or a large value) |
– | The level of correlation between individual claims and various external factors (for example, sensitivity to financial market conditions vs. sensitivity to liability/tort conditions vs. sensitivity to weather events) |
The relative importance of these characteristics varies by the purpose of the user - for example, an excess of loss reinsurer may be primarily concerned about the prevalence of large claims while an asset/liability risk manager may be more focused on the timing of payments and the sensitivity of claims to external conditions. In turn, the exact characteristic that is being focused upon directly affects which type of aggregation produces the most homogeneity - splits by line of business, for example, may produce homogeneous triangles with reference to how quickly claims are reported, but result in heterogeneous triangles with regards to claim size. For example, a triangle produced for personal property would distort claims volatility since some of our personal property contracts are homeowners (frequency) and others are CAT XOL (severity).
Within our frequency, severity and health categories, we have focused on aggregating triangles that are homogeneous with respect to the prevalence of large claims vs. small claims and the volatility of those claims (i.e. the severity category contains business that is more likely to have large claims but where those claims occur intermittently, while the frequency category has a more stable base of attritional claims).
As indicators of the difference in characteristics between the categories, we note the following stats:
Frequency | Severity | |
Average dollars of claims per treaty (US$) | 7,114,857 | 1,172,697 |
Weighted average time to payout (years) | 2.540 | 6.069 |
Standard deviation of paid loss development factor 12-36 mos | 46.5% | 6,761.3% |
Standard deviation of paid loss development factor 36-ultimate | 15.2% | 492.4% |
We expect that the frequency lines would have more stable development patterns given the larger volume and lower level of volatility of the losses that get reported on such business when compared to that classified as severity business. This is indeed the case as shown by the significantly lower standard deviation of the paid loss development factors for our historical frequency business. Regarding the weighted average time to payout ratios, our severity classification includes several casualty classes of business, including casualty clash, E&O and excess casualty type covers that are long-tailed in nature resulting in the severity duration being longer than the corresponding duration for the frequency classification.
Thus, in total, we believe the groupings we have historically used are creating a meaningful aggregation of the business into sufficiently homogeneous categories without obscuring the disclosures with a large amount of insignificant detail. Nonetheless, we continue to analyze our data to determine the most meaningful aggregations for the reader, while also understanding that gains in homogeneity in one dimension may cause additional heterogeneity in another dimension,
2
and/or loss of statistical credibility. As our business strategy and approach evolve, we will consider whether any further disaggregation of these categories would provide more useful information.
Please do not hesitate to contact the undersigned at (345) 749-0205 with any questions or comments regarding any of the foregoing.
Very truly yours,
/s/ Tim Courtis
Tim Courtis
Chief Financial Officer
3