The Intelligent Quarterly from the publishers of The Insurance Insider

Spring 2012

Search archive

Advanced search

Five days of panic-stricken analysis was all it took for financial markets to respond to the 11 March Great Tohoku earthquake, analysis by IQ has found.

Using a sample of Bermudian, Lloyd's and continental reinsurer stocks, the results show that in the immediate aftermath of the quake reinsurer stocks fell and then hit a trough within one to five days - regardless of how long it took the companies to officially announce likely losses.

Additionally, the plight of reinsurers was broadly similar to non-insurance stocks during the period, perhaps indicating a myopic fascination with loss estimates and their impact on the marketplace and valuations.

Five days to trough
Analysis of the nine firms' stock price movements show that the initial period of sell-off panic was actually extremely limited.

All nine firms hit the bottom of their initial post-quake fall within five trading days of the event, regardless of the timing of their expected loss announcements (see table).

Click to enlarge Indeed, the three Lloyd's players in the sample had all troughed within three trading days despite the fact none of them informed the market of their likely losses until considerably later.

For instance, Catlin's post-quake low was just two trading days afterwards despite not giving an official announcement on its loss estimate until 20 April. Indeed, its share price had already recovered to its pre-quake value by 21 March.

In fact, all three of the Lloyd's reinsurers had regained their pre-quake share value within seven trading days. This perhaps reflects the high level of disclosure on probable maximum losses required by Lloyd's syndicates, which allowed analysts to disseminate likely losses and relative exposures within days of the quake.

Meanwhile, it took three trading days for European (re)insurer Scor's share price to hit its lowest post-quake point despite it putting out a fairly un-daunting worst case scenario loss of EUR185mn on the second trading day after (14 March).

The (re)insurer lost 11.4 percent of its value even though its worst-case scenario loss represented just 4 percent of its 2010 year-end shareholder funds.

Continental peers Swiss Re and Munich Re also hit lows within five trading days, although neither released their likely losses until the following week.

A similar story can be seen in the Bermudians in the sample. The three show little correlation between their share price and the timing of their loss announcement.

Obviously, there are bigger forces at work than investor uncertainty over extreme risk exposure and official announcements clarifying insurers' positions.

Firstly, it was widely known that the industry was awash with excess capital despite the dents made by the cat events in New Zealand and Australia during the first quarter. Only an extremely sizeable event would have the power to change that.

In the immediate aftermath it was fairly clear that the earthquake was not the sort of peak-zone catastrophe that could provoke a capital event for the industry. Tokyo, a potential graveyard for reinsurers' exposure, seemed largely unaffected.

Additionally, it was also widely known that international reinsurers would have their ultimate exposure limited by the comparatively low insurance penetration in Japan, the isolationist nature of the primary markets and the earthquake reinsurance programme for dwellings provided by the government.

This was reflected in initial industry loss estimates. Modelling firm AIR Worldwide said on 13 March it expected industry insured losses of just $15bn-$35bn.

Indeed, perhaps the great unknown immediately after the quake, before the attention shifted to possible contingent business interruption losses, was the potential exposure to nuclear contamination cover.

However, this issue was also quickly resolved. Chaucer's specialist Lloyd's nuclear Syndicate 1176 announced on 14 March that legal exclusions under the Japanese Nuclear Act of 1961 would shield reinsurers from liability for any damage arising from a "grave natural disaster of an exceptional nature".

Bottoms up?
It is tempting to consider the insurance industry in isolation and think that relative underwriting performances and loss differences between firms can explain their relative performance - especially after a big catastrophe.

But, tellingly, the story of the reinsurers outlined above closely matches the course of major stock indexes during the period.

The FTSE 100 index of leading London-listed shares and Standard & Poor's (S&P) 500 index both took four trading days to hit their nadir. During this period the FTSE and the S&P indexes retreated by 4.2 and 3 percent respectively.

Click to enlarge The combined market capitalisation of the sample reinsurers fell quicker than these indexes. The nine firms lost circa 9.2 percent ($7.3bn) by 16 March. But even so, using their 10 March closing value as 100 percent, the sample firms clearly match the momentum of their broader corporate environment, if at a slightly larger magnitude in the case of the continental reinsurers (see individual graphs, click to enlarge).

One factor may have been investors gambling on raising rates in the wake of the event - tempering more dramatic falls.

But clearly the risk characteristic of the initial post-quake period followed a similar pattern to broader corporate interests and non-insurance stocks with only indirect exposures to Japanese damage.

A tale of three Bermudians
Few Bermudian (re)insurers would describe the post-quake period as the best of times but, of the sample, executives at Flagstone would have greater reason to describe them as the worst of times.

Flagstone is perhaps the biggest outlier of the group. The Bermudian had tumbled 23.7 percent by its trough on 17 March and was still down 15.7 percent on 30 March after announcing a loss estimate of $80mn-$130mn (or 7.1 to 11.5 percent of year-end shareholder equity).

During this period, of course, the embattled firm was under pressure as Moody's put it under review. The ratings agency cited its outsized losses during Q1 2011, which had been "out of step" with peers and had already caused an estimated 14 percent of its equity to be eroded.

By comparison, Bermudian (re)insurer Arch enjoyed, if not the best of times, then definitely not the worst.

Arch took the unusual step of announcing its loss estimate from the quake on the Monday - and the first trading day - after the event.

Arch said that it expected losses in the range of $35mn-$70mn - far below what was expected from many of its peers. As such, the (re)insurer was given a vote of confidence by the market as its share price rose on Monday 14 March by 0.7 percent, even as the price of many of its contemporaries plummeted.

Indeed, it remained largely stable throughout much of the turmoil after the quake despite its estimate being based on industry-wide insured losses of $6bn to $12bn - far below what cat modelling agencies had predicted at the time and since.

Reputation is king
Perhaps the main story is that reputation is king. Punishment came to those considered to be most at risk of "outsized" losses, not necessarily because of evidence on their exposure or an announcement on their likely losses, but because of a reputation or track record for being overweight on losses compared to their peers.

Risk aversion for the majority appeared broadly in line with initial loss estimates for the industry as a whole and maintained a similar momentum to broader non-insurance linked stocks.