Insurance-Linked Investments: AQR Re: where quakes meet quants
Martin Steward met the reinsurance team established by quantitative asset manager AQR Capital Management
There are plenty of catastrophe risk funds that get called ‘hedge funds’ and plenty of funds of hedge funds that allocate to catastrophe risk funds. There are even multi-strategy hedge funds dabbling in catastrophe bonds. A hedge fund setting up a dedicated reinsurance company is a little different, however – and that is the story of AQR Re.
AQR Capital Management is well-known for its quantitative long-only equity and long/short alternative, and multi-asset risk parity strategies. But in mid-2011 it announced the establishment of AQR Re, an independent Bermuda-based reinsurance affiliate that would raise capital from institutions to invest in natural-catastrophe and other insurance markets.
“We would usually build such a new business ourselves, but we had a long history with a particular team that shared our values,” says founding principal David Kabiller.
It is clear why AQR would want to hire the team that it has. Based in London, the risk sourcing effort is headed by Charlie Vaughan and Rick Montgomerie, both Lloyd’s alumni with 35 and 27 years of reinsurance experience, respectively. They are not insurance-linked securities (ILS) fund managers or analysts, and AQR Re claims this as its differentiator against the majority of ILS funds. It contends that the only real source of capacity and diversification offering keen risk-adjusted returns is traditional reinsurance.
The firm is targeting equity-like risk premiums from roughly one-in-20-year events, and that requires something beyond the peak perils that dominate the securitised market.
“A lot of non-peak will only be bought on a non-traditional basis – and a lot of ILS guys just aren’t set up to do that business,” says Montgomerie.
“But the market is finite – there are only so many insurance companies that buy reinsurance,” adds Vaughan. “You need the relationships to get on the placements. It’s a pretty old-fashioned business.”
Vaughan says that he and Montgomerie have seasoned client relationships with around 160 insurers – from regional independents to multinationals – and that they see 90% of them every year. “They are not just a number on a bit of paper.”
But why go for non-peak perils in the first place? Many argue that because reinsurers have a peril-diversification requirement and non-traditional catastrophe investors do not, the latter enjoy artificially high premiums from non-diversifying peak perils and artificially low premiums from diversifying non-peak perils.
“It’s true that diversifying risks in general do not offer risk premiums as great as hurricane risk,” says AQR Re’s CIO, Andrew Sterge. “But it’s also simplistic: Japan earthquake and Japan typhoon are big peak perils but the risk premiums are relatively low; with non-peak it’s a question of being able to source the very best risks and combine them in an optimised portfolio with the peak perils.”
If you have the relationships to re-insure regional insurers in places like Wyoming and Montana, which have very little peak-peril exposure and actually want to offload non-peak risks to diversify their balance sheets, you can be very well rewarded for risks that add diversification to your own portfolio, Sterge argues.
“Other managers may tell you ‘only go peak’,” agrees Vaughan. “But that’s probably because they can’t access the non-peak risks.”
So why would this team want to work with AQR? And how did they meet?
The key is in Sterge’s unusual career, which began in options and high frequency trading with CooperNeff. It was this quantitative background – Sterge has a PhD in mathematics – that caused his path to cross AQR’s over the years. As CEO at CooperNeff he diversified into catastrophe bonds in 1999, and when he left in 2005 it was the two areas of quant equities and reinsurance that he took to form AJ Sterge Investment Strategies. Vaughan and Montgomerie joined soon after to source risk from London. The business was soon acquired by one of its biggest clients, Magnetar Capital, and the team ran its reinsurance business until Magnetar had to retrench in 2008. That was when conversations with AQR began.
Sterge has never been in a role to build the institutional investor network that AQR has cultivated but just as important is the deep appreciation of the portfolio optimisation expertise the firm deploys.
“We have access to capital from investors who would like this diversification,” says Kabiller. “But we also have a lot of coding, systems, risk and capital markets optimisation expertise, and we think we can make a meaningful contribution to how that applies to reinsurance.”
It is a meaningful challenge. Optimisation based on required levels of compensation for negatively-skewed tail risk is possible – but whereas an entire universe of liquid securities can be analysed all at once and fed into an optimiser to yield desired allocations, in reinsurance optimisation has to be implemented, piece by piece, with a ‘hill-climbing algorithm’.
“At most, you have perhaps 30-50 truly independent risks, and some of those you will not be able to put on in the same size as others,” Sterge explains. “If you are considering a risk that is independent from the rest of the portfolio you would have a lower required risk premium to put that on. But if you were considering a new Gulf hurricane risk, for example, that would increase the tail of your portfolio, so you would require a higher premium to add that risk to the portfolio. Each addition to the portfolio is judged on this marginal utility to the portfolio.”
The marriage of the quake and quant expertise is suitably elegant: Vaughan, Montgomerie and their team will source diversifying risks with keen premiums, and portfolio construction by Sterge and the AQR quants will maximise that diversification benefit.
The initial plan is to provide two or three products with different risk profiles. Reflecting AQR’s recent analysis of Berkshire Hathaway’s use of its insurance float as leverage on a high-quality, low-volatility investment portfolio, the next ambition is to provide a product that takes the reinsurance collateral provided by fund investors, currently held in US Treasuries, and invest it in an AQR high Sharpe-ratio strategy.
“We have the capabilities to marry those two elements and my sense is that this kind of product is still a year or so off,” says Kabiller. “There are a lot of regulatory constraints on what you can invest the collateral into, so getting that balance between quality, liquidity and Sharpe ratio correct will take some more work.”
In other words, expect more innovation from AQR in this resolutely “old fashioned” world of reinsurance.