American International Group Inc. (AIG, Financial) is a well-known global property and casualty insurer paired with a U.S.-dominated life and retirement business.
Given the broad, generalized awareness of AIG and its businesses, I will focus more on its current dynamics and the challenges it has faced in recent years, determining if brighter days are ahead.
To begin, the past several years have been gloomy for AIG, as evidenced by high combined ratios and significant reserve charges in the property and casualty business that have led to subpar returns and the stock trading well below book value. Continued underperformance in the property and casualty business after repeated reserve strengthening and promises of still-to-come higher returns have left investors feeling like Charlie Brown after Lucy pulled away the football, yet again.
With that as a backdrop, why wade into these waters?
P&C reserve risk
In recent years, AIG has taken significant prior-year reserve charges, which have severely hampered returns and brought into question the credibility of its loss reserves. These charges have not only hurt reported results, but have also harmed investor confidence, contributing to significant share price declines. This year alone, AIG is off almost 35% from highs in the mid-60s.
AIG increased prior-year reserves by $3.6 billion in the fourth quarter of 2015. The company followed that up with a fourth-quarter 2016 reserve charge of $5.6 billion. This charge was announced in February of 2017, one month after disclosing a large reserve risk transfer transaction with Berkshire Hathaway (BRK.A, Financial)(BRK.B, Financial).
Then AIG announced yet another reserve charge of $863 million for the third quarter of 2017, shortly after new CEO Brian Duperreault took over in May 2017. These repeated reserve charges have felt like death by a thousand cuts and have severely harmed the company's credibility among investors.
Since we are using book value as an integral of value determination, it is important to have confidence in book value. AIG’s property and casualty reserves are a significant determinant of the quality of its book value. Importantly, sitting here today, the property and casualty reserves have been significantly derisked by the painful strengthening from the previously discussed prior-year reserve charges and the Berkshire Adverse Development Cover (ADC), which provides coverage above a threshold for property and casualty reserves prior to 2015.
Thus, measures of book value are stronger today than before the reserve charges and the Berkshire ADC and provide incremental confidence in our assessment of value. It should also be noted there has actually been a very small net reserve release in the three quarters thus far in 2018 from the property and casualty businesses, a welcome change from the periods discussed above.
P&C underwriting profitability
AIG has also been wracked with subpar returns, even excluding these reserve charges, largely stemming from poor results in its property and casualty businesses, driven by high combined ratios (combined ratios of greater than 100 come from all the costs of writing and servicing insurance policies and claims being more than the premiums received). A return to underwriting profitability (combined ratios less than 100) would raise returns on equity and ultimately lead to more normalized valuations for AIG.
Duperreault has committed to underwriting profitability for 2019. Can the company do it?
A large operation such as AIG cannot be turned on a dime, but Duperreault and his team have taken measurable steps to improve risk selection and underwriting. Several examples of the changes being made were discussed at a recent KBW investor conference (I would encourage anyone interested in AIG as a potential investment to listen to the archived webcast, which can be found on AIG’s Investor Relations site). Of note, there is now much more discussion about the particulars of underwriting policies, including risk limits and attachment points. In years past, management discussions centered more broadly on the levels of written premiums alongside expense saves and buyback plans. While expense saves and buybacks at these levels are good for the business, the relative emphasis away from thoughtful and disciplined underwriting was likely taken too far.
I will share one example that drives home both the new approach and the old challenges. AIG has a surplus lines unit for business where the risk is not admitted into the regular market and where rapid decisions are needed. The historical process at AIG was too slow and they would often end up with the worst kind of surplus business (essentially, the leftover surplus business or the leftovers of the leftovers) because they were operating at a speed where only the worst risks remained when they were ready to move. The slower internal process (running the surplus lines through regular channels) created its own adverse selection of risks.
More broadly, Duperreault spoke of the need to heighten focus on risk selection and attach points rather than just pricing, standing in contrast to the idea that all risks can be accounted for with enough pricing. He spoke of the need to drive up attach points to stay above loss levels as an indirect way of getting more “pricing,” even if the rate does not change. Additionally, AIG’s new Chief Actuary Mark Lyons spoke on the third-quarter call about how the company historically would use higher capacity limits to compete, thereby increasing potential losses in adverse scenarios. The implication from these and other discussions is that this new team has a deeper understanding of how to structure a portfolio of risks to produce better underwriting results.
One particular stream of comments during the KBW conference struck me as a corollary to value investing: when speaking about risk selection, Duperreault pushed back against the ability to unilaterally move pricing levels to accommodate for risk, stating that the market largely dictates pricing. The key then becomes choosing the right risks given the level of market pricing available. In investing parlance, the best investing risks to take are largely dependent on the prevailing stock price at the time of investment. You cannot move the stock price, but you can choose the investment where the value received is higher than the price paid. For AIG, the key is taking the right risks at the right attach points and policy limits given the level of pricing the market will bear.
The broader and more important point to take home from these recent talks and presentations is that these concepts are now openly and fully discussed by AIG. It is clear that the new leader has a background firmly entrenched in insurance businesses, including a long sting at AIG earlier in his career, whereas Peter Hancock came from a banking background. This is an important change in my view, but one where the results may not be fully seen for some time to come. Given the nature of insurance contracts and their renewal patterns, improved results are not likely to be seen until 2020 or 2021, when the underwriting decisions of today will start to bear fruit in reported results. There should be early indicators in 2019, when the team’s commitment to underwriting profitability as a first step in its improvement can be tested.
Recent catastrophe losses
Given our long time horizon, we try to not spend too much time obsessing over recent news, much preferring to think about the long-term issues facing businesses. However, AIG’s recent catatastrope losses have been in the news and they do raise an important issue. AIG reported reasonably large losses in the third quarter from typhoons in Japan and the two recent U.S. hurricanes. On the KBW call this past summer, Duperreault mentioned they had bought more catatstrophe reinsurance for this year, not wanting to repeat the large losses of 2017. While the third-quarter headline catatstrophe losses may seem to call his credibility into question, we see a different perspective and would assert he has gained incremental credibility.
Reading the Oct. 18 release about the catastrophe losses, there is a sentence that says with the North American reinsurance in place, the catastrophe exposure for the remainder of 2018 remains limited given previously purchased reinsurance. This would imply much lower catastrophe losses for 2018 versus 2017 even if there are significant events still to come. Thus, this bout of bad news actually served to bring more, rather than less, credibility to AIG’s management team.
There has been considerable discussion by the AIG team about the work being done to dampen the volatility of underwriting results in 2019 and beyond. The company has reduced gross and net limits, increased attachment points and expanded the reinsurance program. They are also working on an expanded Japan catastrophe reinsurance program for 2019 to complement the expanded North America catastrophe reinsurance. The results of new volatility dampening actions will not be seen until 2019 and 2020, but management’s focus appears to be narrowed in on more consistent and more profitable underwriting.
Price to value
AIG’s book value per share at the end of the third quarter was $66.23 and its book value per share, excluding accumulated other comprehensive income, was $66.83 (AOCI moves up and down in value with changes in the value of its bond portfolio that is used to cover its long-dated liabilities). Book value per share, excluding both AOCI and its deferred tax assets, was $55.58. With AIG trading around $44, there is a clear margin of safety using even the most conservative measures of value. Even a return to book value would yield strong returns from current price levels with strong downside protection. Many competitors to AIG trade in the neighborhood of 1.5 times book value.
Furthering our narrative on value, AIG’s consensus earnings for 2019 are $5.11 per share, leaving shares valued around 8.5 times coming-year earnings with shares trading near $44. This level of near-term earnings is supported by management’s statements that 9% returns on adjusted equity (equity less AOCI and DTA) can be achieved in the relative short term.
We can also consider AIG in this way: it is a parent holding company that sits above and owns its various insurance businesses. For this year, even with the elevated catastrophe losses, those insurance businesses will distribute roughly $6 billion to the parent company. With a market cap below $40 billion, annualized cash flow to the parent provides further evidence of good value at today’s price levels.
AIG moving to a price of $65, which would still be less than current book value, would bring strong returns from today’s levels. Valuations closer to historical and current competitor levels would yield even stronger returns.
Furthering on the thesis, AIG’s Troubled Asset Relief Program warrants trade around $9. These warrants expire in January 2021 and have a strike price that declines for quarterly dividends above a threshold. The number of shares received also increases each quarter a dividend is paid above the threshold. The original strike price was $45 with a one share per warrant share count factor. Today, the strike has come down to $43.6582 and the share count factor is 1.032. By expiration, I estimate the strike could reach $42.50 and the share count factor could reach 1.06. (The warrant prospectus details the formula for these quarterly changes and should be read by any interested in the warrants).
If AIG moves to a price of $65, the warrants could be worth about $24, well more than double current levels. As book value per share increases to roughly $75 at expiry in early 2021, the warrants could be worth $34, more than three times current levels, assuming a valuation around book value. Valuations closer to historical and current competitor levels would yield even stronger returns.
The time to warrant expiry is much shorter than my preferred investment horizon, which brings timing risk to the forefront and make the warrants only suitable for those willing to take such heightened timing risks. AIG is likely to improve its business over time and its valuation should follow, but the improvements and market recognition of such would need to occur between now and January 2021 for the warrant thesis to bear fruit. I would also note the AIG warrants today carry a premium which is in contrast to some of the recently matured TARP warrants, such as those for JPMorgan (JPM, Financial). This premium level also increases the incremental risk of the warrants relative to the common shares.
Risks
There are two sources of risk to discuss: AIG itself and risks particular to the warrants.
We have talked about historical reserve charges destroying shareholder value and any future large reserve charges would also harm AIG’s value. In addition, the company has not consistently posted appropriate returns on equity, which will be necessary for a return to a more normalized valuation. Any setbacks in the property and casualty business are a looming risk. Additionally, while incremental increases in interest rates should be a positive for AIG, significant and sudden interest rate movements could hurt the business. Further, rather than recounting all the relevant business risk, please review their most recent 10K for a full rundown of such risks.
As discussed above, the warrants present an added degree of risk in that the normalization of value needs to occur before January 2021 for the warrants to see commensurate increases in value. This is a comparatively short time horizon for my style of investing, so I have balanced my approach by owning mostly common shares with a much smaller allocation to the warrants. I am confident AIG can regain higher returns and valuations, but I am less certain of when that will occur.
The shorter time horizon also introduces more general market risk as market downdrafts are likely to hamper the return to more normalized valuations. This is not a great concern with our more customary 10- to 20-year horizons, but is a concern specific to the warrants.
Disclosure: The opinions in this document are for informational and educational purposes only and should not be construed as a recommendation to buy or sell the stocks mentioned or to solicit transactions or clients. Past performance of the companies discussed may not continue and the companies may not achieve the earnings growth as predicted. The information in this document is believed to be accurate, but under no circumstances should a person act upon the information contained within. We do not recommend that anyone act upon any investment information without first consulting an investment adviser as to the suitability of such investments for his specific situation. A comprehensive due diligence effort is recommended.
Read more here.
