Evolution of the Property-Catastrophe Reinsurance Market
This year’s 8 percent Guy Carpenter World ROL Index increase differs profoundly from the 65 percent surge that followed Hurricane Andrew and the 24 percent hike following the terror attacks of September 11, 2001 in the United States. Even after losing 18 percent of its aggregate capital following the 2008 financial catastrophe, reinsurers were unable to push for the high rates that some expected. The evolution of the reinsurance industry over the past two decades suggests that carriers have become much more adept at managing risk and capital, making it easier to absorb shock losses and manage the cost to transfer risk.
Following Hurricane Andrew in 1992 (loss of USD23 billion), ROLs increased by 65 percent at the 1993 renewals. The rate increases can be decomposed into two major parts: the pure premium increase and the impact of a decline in reinsurer capital.
The first part was the “pure premium increase,” as reinsurers saw the claims-driven costs increasing substantially because of the storm. The increased potential for insured losses resulted in premium increases in the primary market, which flowed up to the reinsurance market, increasing cedents’ costs to transfer risk. Quite simply, reinsurers believed that the average size of a loss that could come from a major tropical cyclone had increased.
The second part of the rate increase following Hurricane Andrew was the result of price pressures from a reduction in capital across the reinsurance industry following the catastrophic loss. The inability to replenish capital quickly drove ROLs higher: the reinsurance industry was a relatively closed financial system at the time. The capital already in the reinsurance system was essentially fixed. With scarce capacity, prices rose sharply, as the market distributed the scarce capital to the highest bidders.
Following Hurricane Andrew, capital did flow into the reinsurance industry in the form of the newly created mono-line catastrophe writers (i.e., the “Bermuda monoline cats”), but because the reinsurance system was closed, it took a while. The first reinsurance start-up to open its doors did so well past the January 1, 1993 renewal date.
The next step in the industry’s evolution came with the terror attacks of September 11, 2001 in the United States. At the subsequent renewal, ROLs increased by an average of 24 percent. The pure premium increase was the dominant factor, based on the additional cost expected for covering the terrorism peril. Before this event occurred, reinsurers had charged minimal amounts for terror losses (if anything at all). Of course, it is difficult to isolate the implications of the terror attacks on reinsurance rates, as price increases were also affected by the asset-side losses — and broader financial implications — of the decimation of the early “dotcom economy” — from the initial NASDAQ drop in March 2000 through the collapse of Enron (and other large companies) in late 2001 and early 2002.
The lower ROL increase at the renewals following the terror attacks resulted largely from a more open reinsurance system. The advent of catastrophe bonds in the late 1990s provided an alternative source of capacity, which, though small relative to traditional reinsurance, showed that capital could enter the industry more easily. Also, sidecars began to gain some prominence with Olympus Re, which was established by White Mountains Insurance Group to take a quota share of Folksamerica Re. This was only 3 percent of the USD16 billion raised in late 2001 and 2002, and alternative capacity (which includes both catastrophe bonds and sidecars) accounted for 14 percent. Despite the increasing openness, the industry still received most of its replenishment capital from traditional equity and new underwriters - such as Axis, Arch, and Endurance.
An industry defined by the transformative effects of shock losses changed again with Hurricanes Katrina, Rita, and Wilma and the 2006 reinsurance renewal — as combined insured losses reached USD60.5 billion. Despite the extent of the damage, these storms did not cause substantial losses of capital, as the rest of the property and casualty (P&C) insurance industry was highly profitable. The insurance industry as a whole recorded a combined ratio of 100.7 and a rate of return of 10 percent. Further, the reinsurance industry was no longer a closed system — external capital was readily available. More than USD35 billion flowed into the industry through a variety of instruments, including start-ups, catastrophe bonds, and sidecars.
Among the reasons for the relatively rapid replenishment was the rise of the alternative investment sector. Hedge funds in particular were attracted to the outsized returns available from the sidecars issued following the 2005 hurricanes. Of the capital raised in late 2005 and 2006, 20 percent came from catastrophe bonds and 13 percent from sidecars. The proportion of capital raised through alternative sources more than doubled from 2002 to 2006. With one-third of the replacement capital coming from non-traditional means, the reinsurance system was far more open than it had been only five years earlier … and was almost unrecognizable (in this regard) compared to the industry in 1993.
Unlike the other industry-changing shocks of the past 20 years, the events of 2008 came mostly from the financial community — and at the same time as a major property catastrophe (i.e., Hurricanes Gustav and Ike). The financial crisis had been developing since February 2007 and erupted in September 2008, dwarfing the economic damage caused by the dotcom collapse. On the same weekend, Hurricane Ike, one of the most expensive hurricanes on record, tore through the Gulf of Mexico, stretching much farther inland than expected.
The events of 2000 and 2001 were quite different from those of 2008 (with effects stretching into 2009). The dotcom collapse was a financial event of a much smaller order than the financial catastrophe of 2008 which threatened the global economic system. Though the (re)insurance industry has emerged from the financial catastrophe — certainly more successfully than other sectors of the financial services industry — it did sustain significant investment losses across all asset classes, particularly fixed income. The dotcom fallout was more pronounced for speculative investments, rather than the conservative portfolios typically held by insurers and reinsurers.
Further, the property-catastrophe shock of 2001 was different in scope and impact than Hurricane Ike. The latter was a large storm, though not a key event for the global reinsurance industry. The terror attacks, on the other hand, shocked the industry, because the mindset had been that terror events were more a threat to persons, with limited potential for large economic loss.
Because of the unique nature of the financial catastrophe, the pure premium impact has not been substantial, as evidenced by the contained average ROL increase around the world. Rather, the losses were caused by the availability of capital relative to the amounts lost. Though there have been signs that efforts to raise capital are becoming more successful (especially from the second quarter onward this year), it will take a considerable amount of time for carriers to return to the robust positions of late 2007 and early 2008.
Not all price changes were related to capital; some reinsurers did charge higher rates for areas hit by Hurricane Ike, particularly for situations where retrocession support had dried up. But, the losses related to the storm were not transformational. The capital implications resulted from the concurrent financial catastrophe. Most sources of outside capital closed, as banks and alternative investment institutions were hit more severely than reinsurers. This effectively returned the industry to a closed financial system to a degree not seen since the 1980s.
As financial markets around the world continue to stabilize — and eventually return to normal — the impact of the loss of capital will be diminished. The reinsurance system will begin to open again, with reinsurers able to replenish their balance sheets more quickly through renewed access to sources of capital outside the industry (including insurance-linked securities). As this happens, the primary driver of property-catastrophe rate changes will be the pure premium factor in the wake of a disaster. The net result is likely to be a less volatile reinsurance market.
Economic Losses Point to Reinsurance Growth Potential
The transformational losses that have pushed the evolution of the reinsurance industry generally have been limited to the availability of capital relative to a fairly consistent marketplace. Reinsurers see the opportunity for rate increases following a catastrophe event or when capital is in short supply. This allows for little flexibility: risk-bearers can’t take their fates into their own hands. Yet, there is another measure of growth potential - and it is one that reinsurers can use to increase top-line and bottom-line performance.
Expanding the insurance and reinsurance systems to areas not currently covered suggests a significant increase in the potential size of the reinsurance industry. By distributing risk more widely around the world, reinsurers can identify new coverage opportunities, write more business, and manage volatility. So, it is possible to increase the size of the marketplace while managing risk and capital more effectively.
Traditionally, analyses of reinsurance industry growth potential have focused on insured losses. This approach may seem sensible, but it has significant weaknesses. The definition of insured loss differs from country to country and over time. Hence, insured losses can lead to a distorted picture of the true loss history. For example, in the United States, flood risk is insured by the federal government, leading insurance studies to neglect losses from this peril. In emerging markets — such as China and India — insured losses tend to be relatively low. This tends to give rise to the potential for small sample error: projections based on small samples can be egregiously inaccurate.
Economic losses, on the other hand, cover a wider range. By measuring all damages — insured and not — it is possible to gauge losses not currently insured but which could be. Effectively, this approach includes in the reinsurance industry’s revenue potential the impact of primary market growth into new regions and risks. The result is a more accurate view of the possible size and composition of the reinsurance industry.
From 1950 through the present, economic losses from natural disasters have grown rapidly, suggesting that there is plenty of growth potential for the reinsurance industry. Coastal population growth is largely responsible for this change over the past seven decades, and it has been compounded by increased property values in exposed areas (e.g., Florida and California). In emerging economies, rapid growth in assets also led to higher exposures.
Economic losses from 1990 to 1999 were 14 times higher than from 1950 to 1959. The rate of economic loss fell 20.2 percent from the last decade to the current one but is still 12 times higher than in the 1950s. The recent decline is probably an anomaly and is likely to disappear if we include man-made catastrophe events, such as terrorism.
The history of the property-catastrophe reinsurance industry indicates that it has the potential to be a fast growing segment of the global economy. This trend is likely to continue in this direction as emerging economies transfer more risk to insurance and reinsurance companies. But, one cannot expect this growth to follow a smooth trajectory, since property catastrophes do not follow a stable path. Hence, long-term players in the market will need larger amounts of capital to withstand the wide swings in catastrophe losses that are likely to occur.