Some of the drivers which will influence the market in 2019 are already apparent. As referenced earlier, changing dynamics in the casualty space, especially in the US, are shaping up to be a prominent feature of forward earnings. It is difficult to overstate the influence of the global casualty market on the reinsurance sector. It has been a key catalyst in nearly all past market turns. The difficulties posed by estimating total ultimate losses for long-tail business mean capital levels in the reinsurance sector become uncertain when reserves, which can represent multiples of annual earned premiums and equity, begin to appear deficient – even at the margin.
The potential implications of a tightening casualty market are therefore clear. After years of largely favourable conditions (e.g. a benign inflationary environment and historically low loss experiences), increasing claims severity and social inflation are now hurting carriers and point to a market in transition. The 1 January 2019 renewal reinforced this sentiment, with a floor established for rates and pockets of firming starting to surface in a number of business lines.
Whilst the casualty market currently remains orderly, and cedents with solid reinsurer relationships can still expect to secure competitive terms, deteriorating results are expected to take their toll as 2019 progresses, especially as reserve redundancies used in recent years to bolster results have now been largely exhausted (see Figure 6). Building inflationary pressures are only going to compound the situation.
Figure 6: Calendar Year Reserve Development by Quarter for Top 30 Global P&C Carriers Versus Accident Year Reserve Experience – 1998 to Q3 2018 (Source: JLT Re)
The property market has also undergone important change over the last 18 months after one of the largest ever capital raises post-HIM prevented the type of pricing correction that had followed previous large loss years. In addition, it quickly became clear that the pricing levels recorded at the turn of 2018 would act more as a ceiling than a floor as price momentum slowed through the course of the year, especially at the 1 June renewal.
That significant and broad rate increases did not materialise in the property market following record losses represented an important break with the past. Capitalisation levels were crucial to this outcome.7
After years of strong capital growth and lacklustre premiums, the reinsurance market last year had more capital relative to risk than at any time in living memory, even after the record losses of 2017. The appetite and ability of ILS markets to reload so smoothly in late 2017 through to the end of the first half of 2018 suggested further significant capital growth would persist into 2019. impaired earnings and even absorbed some of the excess capital in the sector.
Figure 7 shows that an additional USD 80 billion of insured catastrophe losses were sustained in 2018, the fourth highest ever in real terms. These above-average costs came fast on the heels of the record-breaking loss-year of 2017. Losses for both years combined are the largest ever to impact the (re)insurance sector over a two-year period, at USD 230 billion. In fact, more than USD 200 billion of insured catastrophe losses have occurred over the last 18 months or so (from HIM onwards), with reinsurers and retrocessionaires covering a large portion of these costs.
Figure 7: Global Insured Inflation-Adjusted Catastrophe Losses – 1970 to 2018E (Source: JLT Re, Swiss Re)
A recent JLT Re study showed that capital has been, and remains, the dominant reinsurance pricing driver: JLT Re Viewpoint – Rethinking the reinsurance cycle,
Overall loss trends are clearly up, and the diverse and dispersed nature of events over the last two years shows that carriers must prepare for higher levels of catastrophe losses in future. Wildfires especially are coming under closer scrutiny after losses in California in 2017 and 2018 spiralled to become significant reinsurance and retrocession events. These outsized losses are not commensurate with a view of risk that has traditionally considered wildfire to be an attritional (rather than a catastrophic or peak) peril. Underlying risk assumptions for multi-peril, aggregate products are now likely to be reassessed to reflect more closely the contribution wildfires and other ‘attritional’ perils can make to overall expected losses.
Figure 8: Claims Development for Costly North Atlantic Hurricanes – 2000 to 2017 (Source: JLT Re, PCS)
Another unexpected ‘event’ to hit reinsurers in 2018 was loss creep associated with HIM, especially Hurricane Irma. Figure 8 shows that losses for Irma, and Harvey to a lesser extent, have developed adversely for longer than normal (beyond 400 days after landfall), due in large part to complex and unmodelled loss components such as loss adjustment expenses and assignments of benefits. This prompted a number of reinsurers and ILS markets to make significant revisions to their initial loss estimates, adding to loss burdens for the year.
Accompanied by further sizeable insured catastrophe losses in 2018, total catastrophe costs for the calendar year were in reality significantly higher than the headline figure of USD 80 billion. Taken in aggregate, creep from HIM was one of the more significant losses to impact the sector in 2018, with reinsurance carriers incurring the bulk of the marginal costs. This manifested itself during the renewal, leading to capacity constraints for certain high-loss lines of business, such as retrocession.
Loss development should again be monitored closely in 2019, especially as several prominent 2018 events occurred outside peak zones and not all losses fell within modelled expectations. Changes to certain commercial catastrophe models may follow, given that performance appears again to have fallen short of the levels desired by the market.
Loss experience is just one of several factors likely to create pricing pressures this year. Moderating ILS capital inflows, rising loss cost trends and reduced capacity at Lloyd’s are also likely to resonate in 2019. However, historically high levels of excess capital will continue to weigh against these dynamics, and, as mentioned earlier, this has long been the dominant reinsurance pricing driver.
Important differences between 2017 and 2018 have nevertheless emerged as new capital inflows have slowed. Whilst the market continues to be well capitalised, as much as USD 20 billion of alternative capital was trapped and not deployable at the 1 January 2019 renewal. Crucially, Figures 9 and 10 show that 2018 did not come close to replicating 2017’s ‘great capital reload’. Indeed, whilst approximately
USD 8 billion of new capital entered the sector in the fourth quarter of 2017, less than half (some USD 3.5 billion, when excluding capital dedicated to mortgage transactions) was raised during the same period of 2018. As a result, there is much speculation about investor appetite going into 2019 and what it means for capacity availability and pricing.
Figure 9: Announced New Reinsurance Capital – Q4 2017 (Source: JLT Re)
In order to assess investor sentiment, it is helpful to understand why such a large and growing number of institutional fund managers has been attracted to collateralised and securitised reinsurance over the last decade. Investors have long measured fund performance against two key metrics. The first is relative return or alpha. The second is volatility, as measured by standard deviation which is itself a function of correlation. The most common measurement of investment performance is the Sharpe Ratio which is, in the simplest terms, a measure of return divided by volatility.9
The effect of catastrophe insurance-linked investments on portfolios can be highly positive, especially if yields are strong and correlations are expected to be low. Correlations are thought to be low due to the theory that large, insured catastrophe losses should be uncorrelated, or at least less correlated to high-yield, equities, commodities and real estate markets, than these investments are to one another.
Appetite for catastrophe bonds and collateralised retrocession increased significantly following the financial crisis, which was an alpha-destroying, high correlation event for most managed funds. The effect of insurance-linked investments on diversified portfolios since then has been largely positive. Low to average losses and relatively high yields meant that they enhanced Sharpe Ratios from 2012 through to 2016. By contrast, 2017’s and 2018’s combined record-breaking losses coincided with large impairments in most other asset classes, especially in 2018. As has been discussed, large, post-loss rate increases did not materialise, dampening some investors’appetite for additional insurance-linked investments. Figure 11 on page 28 shows US catastrophe bond index prices compared to other asset classes in 2017 and 2018.
As 2018 progressed and investors surveyed the damage, a new picture began to emerge. With lower expected net yields, and what appeared to be higher correlations to other asset classes, the benevolent effect of catastrophe insurance-linked investments on portfolios, whilst still positive, abated in the minds of some investors.
Figure 10: Announced New Reinsurance Capital – Q4 20188 (Source: JLT Re)
Figure 11: Is Everything Correlated in the Tail? (Source: JLT Re)
Figure 12 shows the evolution of the effect on Sharpe Ratios of progressively higher allocations to catastrophe insurance-linked investments.
This goes some way to explaining relatively lower inflows of new investment capital into the sector following 2018’s catastrophes when compared to 2017. Whilst it may be tempting to conclude that investor appetite for ILS and collateralised vehicles has abated, it is important to remember at least three counter-points to this argument.
Firstly, third-party capital investments as a percentage of dedicated reinsurance capital remain at or near all-time highs. There has been no largescale investor exit following the largest real-terms insured catastrophe losses ever. On the contrary, additional new capital has entered, albeit not in the same quantities as a year ago to date. Secondly, Correlation and causation can be two very different things, and investors understand this. Finally, some of the decrease in ILS pricing is due to the current unavailability of fungible capital from other asset classes given poor performance elsewhere in portfolios. This has probably led some ILS investors to liquidate positions in the secondary market in order to fund new vehicles, thereby exacerbating the correlation with other asset classes.
It should also be remembered that loss-affected retrocession ROLs have risen by double-digits for two years running. This, combined with lower secondary market pricing, means that yields on newly formed investment vehicles should rise in line with rate increases. For the patient investor with capital to deploy, conditions may, from some perspectives, be improving.
Figure 12: Effect of Collateralised Retrocession and ILS on a Hypothetical Investment Portfolio as Measured by Sharpe Ratio – 2017 to 201810 (Source: JLT Re)
10 See appendix for hypothetical portfolio allocations.
Figure 13: Hypothetical Portfolio Allocation (Source: JLT Re)