Down front if you want to make your way down. If you haven't met me before, I'm Jay Cohen. I'm the Senior Property Casualty Analyst at BofA Merrill. My pleasure to have Peter Hancock, President and Chief Executive Officer of AIG, as our next presenter. AIG hasn't presented at our conference in many years. Just the timing didn't work out. We're fortunate to have Peter here. Timing is interesting as well, given a pretty noticeable fourth quarter. Peter joined the company, was it 2010?
Right.
First head of risk management, but really over the past seven years has tried to put his imprint on the company, made a fantastic amount of changes in the organization. In the fourth quarter, obviously there was a sizable reserve charge. There was a big reinsurance cover. We'll talk about that. As you go through their disclosure, it actually does add some insight into how we look at the company. We do have a lot of ground to cover. I want to start, Peter, with kind of maybe a big picture question, just to give you a chance to chat a bit about 2016. I think we tend to look at one quarter, but you've been on this journey of changing this company. How do you view 2016 in that lens?
Also just thinking about 2017 for a bit.
Jay, thank you very much for having me. I couldn't be happier about the timing, because obviously yesterday was a very big day for the company, a lot of news, and a lot of market reaction to it. I'd like to use this as an opportunity to keep things in proper perspective, because as you say, 2016 was a year with a lot of change and a lot of accomplishments. While I'm not happy with the fourth quarter, I'm very happy with the total progress of the year, because it was a very pivotal year for the company. I made a very significant change to my management team at the beginning of 2016, completely new lineup.
They have executed our strategic plan that we laid out at the beginning of the year in a way that I find very comforting in terms of confidence in our future earnings outlook. Also cleaning up a lot of the risk that had hung over the company, whether it's in the legacy portfolio, which we have shrunk quite dramatically, freeing up over $10 billion of cash flow last year through divestitures and reinsurance transactions, or radical shifts in the mix of business. I think Rob Schimek, who has been running commercial for the last year, can take a lot of credit for corrective actions and really rebalancing the book from a heavy overdependence on the casualty business, the U.S. casualty business, to a more balanced mix of business. Our mix of business is better positioned today than a year ago by quite a margin.
On the expense side, over-accomplishing on the expense targets we set, taking $1 billion out of expense in the last year. That's the second year in a row of beating expense projections, and with continued momentum on the expense side, which has given us, frankly, the courage to give up top line where we haven't received pricing. The prior year development that we announced is indicative of quite how tough the pricing environment is, especially in the casualty lines and in certain other lines as well, like the E&S property space. I think that we're really reshaping the portfolio to a more sustainable mix around the clients that value what we do most. This is a very important moment, I think, in the company's history in terms of putting the past behind us.
If you look back at our long history, there are moments where we were by far the dominant player in certain sectors. U.S. casualty, I'll just use as an example, 10 years ago, we were writing about $15 billion of net premium written. The year to come will be about two and a half. Very dramatic change, but the reserves on that don't go away. The adverse development cover that we entered into with Berkshire three weeks ago was a critical way to really truncate the reserve risk on that long history and rightsize it relative to the new business we're writing. That's really been a big story of AIG in general over the last seven years, which is really narrowing its focus on where we can win. The client sectors, the geographies, and the product sets where we have real comparative advantage.
2017, what are the key priorities for you and your team?
For 2017, we really want to continue executing on this plan. Continued expense discipline, capital return. We've reaffirmed our commitment on capital. That's a $25 billion target. With the important caveat, which I've made consistently, including a year ago and prior to that we view the view of the rating agencies and state regulators as paramount because our clients need to be very comfortable with our claims paying ability. The holding company, with over $8.5 billion of liquidity, is a very strong backstop to the subsidiaries. The subsidiaries now have the adverse development cover, which really buttresses their reserve risk. We need to stabilize the rating outlook, and I'm delighted that Moody's published last night. I think that we have increasing confidence from all of the stakeholders that we're prudently managing capital. That's one priority.
Expenses, executing an improvement in the mix of business. As you see the changes on the net premium written in 2016, you'll see it earn into net premiums earned. I think that that is what gave us the confidence to give guidance on a 9.5% ROE on the core portfolio, which is what you need to think about as your longer-term Earnings engine as opposed to the legacy which we continue to divest. We made some good progress on divesting legacy assets and reinsurance of some of the old blocks of the life business in the fourth quarter. Continued divestitures of the legacy and continued operating improvements in the core portfolio.
Got it. I want to jump into the reserve action. I think one of the bigger issues people had, and it certainly was an issue for me, is that while the charge was five and a half billion, 1.3 of that charge was for 2015. It added six points to your commercial loss ratio. To me, that says these guys have a real problem in pricing their current business. Yet you're coming out with guidance for 2017. How do we have any confidence that you guys can do that given that you just added so much to the reserves after one year after taking a charge as well?
Obviously, especially for longer tail lines, I think if you compare us to our peers, we have about 50% longer duration of claims half-life than others. Our mix of business given our participation in the excess layers means that you have a longer emergence, means that we were taking very early action on very green accident years by doing that, which we think was a prudent thing to do. With the ADC on a net basis, it really emphasizes that we're sharing the incremental risk going forward, 80/20 with Berkshire. We see emerging trends that are worrisome. I think that industry wide, others may not be recognizing them as promptly, but we certainly see them in the increase in very large verdicts on personal injuries and obviously the increase in trucking accidents in terms of frequency and severity.
Again, but you're looking at 2017, you've got a price business. Mix has changed a bit, but there's still some of this longer tail business. Have you changed something in your methodology where you're going to get it right this time, you'd hope?
Well, I think the biggest change is dramatically shrinking the amount of this business we're writing. Year on year, 2015 to 2016, a 35% decline in this business. Dramatic reduction in our exposure going forward to getting that precisely right. Of course, yes, we are refining our risk tools to better segment. I'd say that we have been better at sector judgments than we have at micro risk selection and pricing, but we're getting better at that. Remember, this is not a middle market or personal lines business where statistical techniques can be optimized very quickly. This is larger limit and more manuscripted policies and with a longer emergence where the cycle of refinement of risk selection takes time. The sector bet we made was very clear five years ago, much less U.S. casualty.
We've been shrinking it year after year quite dramatically in the last 12 months, that's what gives us confidence about the overall profitability picture. Casualty's gone from 40% to 20% of our U.S. commercial business. It's just a much more modest bet, there's less at stake. We're obviously also working with our reinsurers, Swiss Re in particular. There's a quota share with us which we entered into the beginning of 2016, comes up for renewal at the end of this year. We've been in conversations with them as recently as this week in the light of the prior year development, we're very confident that that will renew on favorable terms because they have a good line of sight in all the steps we've been taking.
One of my questions was Swiss Re. You think even with the actions you've taken that renewing that contract will happen?
I think especially because of the actions we've taken. I think they have followed very closely the underwriting decisions we've made over the last year. They have contributed their tools working with us. They've seen the dramatic mix shift that the commercial team led by Rob Schimek have done over the last 12 months, and they feel confident in the mix of business going forward, so that they're shoulder to shoulder with us with a very attractive ceding commission.
You had a margin improvement plan for 2016 and 2017, and you gave us pretty good guidance on that. Obviously, the starting point's higher, but the improvement now you're looking for in 2017 seems to be bigger than the initial improvement you were looking for in 2017. To get there, do you have to take other actions that you didn't contemplate earlier? Or were the actions you took last year really the drivers of what will be a more dramatic improvement?
A lot of it are actions that have already taken place. I'm not saying all of it. Obviously there's business to be renewed, but as you know, the lag between written and earned means that we have a pretty good line of sight on the year to come. I think that the mix of business has been quite dramatic. I think where we've obviously seen continued pricing pressure is in the property space and especially on the E&S. We will continue to cut back that. We've also dramatically increased our cat reinsurance purchases so that we've reduced our AAL. Unlike many others, when we look at the property ROE numbers, we normalize it with a cat load that is a function of what reinsurance we do.
Again, having done the one renewal at very favorable terms with a low attachment point, we feel more comfortable with our projections on the property. There is a significant property improvement year-on-year that's embedded in our assumptions. The other thing we're very confident about is on the expense side. A lot of the focus a year ago was the four point improvement in loss ratio that you alluded to, which we did achieve, albeit from a higher starting point in hindsight. We continue to see loss ratio improvement in the forecast. The expense ratio also has contributed to the combined ratio improvement, and that's something where I've been very proud of the way the team has been able to embrace automation, rationalization of business processes, and removing redundant process.
On that topic, arguably you may have to shrink your premiums more in 2017. Will you have to take additional action on the expenses to at least maintain or defend that expense ratio in addition to what you may have contemplated earlier?
We already have. As you can imagine with a company of our scale, there's a lot of inertia. We started on the cost direction the day I became CEO. It didn't show up in my first year because of the inertia. There was a real pivot in the expense trajectory of the firm that coincided with my arrival. It was a strategic judgment that we would have a much easier time making the tough sector decisions from the risk selection point of view if we had no pressure to write business to defend at an expense ratio.
Getting expense momentum going in the right direction has given us the courage to walk away from underpriced business and to focus our resources on investing in future technology improvements because we're maintaining investments in infrastructure, because that's something which we continue to see a need for if we're going to remain competitive.
It's not like you have to take some additional major restructuring action to achieve it. You've already got those plans in place.
It's very much so. It's really executing plans that have already been initiated.
On your reserving track record, which obviously is not great, the interesting thing from us is that it's so different from others out there, right? Some of your major competitors continue to release reserves, you're adding to reserves. My question to you, and I don't think you can quantify it, but I want your thoughts on it. How much of this is a business mix issue? Because that is part of it, but some of it has to be poor execution and whether it's bad systems, too much focus on growth historically. How much is it self-inflicted versus just your business mix?
A lot of different dimensions. I think that our reserving has improved methodologically tremendously over the last few years. It's much more transparent. It's much more bottom up. I think that it creates much more accountability than it ever did. There's still lots of room for improvement. I don't want to claim that we're at the destination. I think that the mix is an issue. We have a lot more high attachment point business where claims don't come until quite late after the accident year. Reacting to that data earlier would be better. I think that that's an important change is to use more industry-wide data on the bottom ground up claims trends rather than waiting for it to hit the attachment point. We've done that. I think resegmentation. I think commercial auto has plagued not only us but the industry.
It occurred in over 25 different reserving segments in our historical reserving approach. The way we've resegmented is to look at how commercial auto as a whole is trending, and that's given us, I think, greater confidence about our reserving. I think we continue to invest in the data, tightening up the linkage between claims, actuarial, and underwriting so we have a faster feedback loop. We made a major centralization of claims under my predecessor. I've gone to a more decentralized model so that the claims and actuarial underwriting are more closely coordinated with each other, which I think has reduced the time lag for reserving judgments. The big difference versus peers is the one I alluded to earlier, the massive downsizing in the long tail business.
Up until the ADC cover we did with Berkshire, we've been reserving for a huge back book of the last 40 years, and you're comparing that to a relatively modest size of new business that we're writing. While most of our peers have been relatively stable in size or growing, where any kind of reserving adjustments were relatively modest compared to the size of their new business. I think that we've right-sized the balance between the old and the new, and that's why I view this as a big pivot point for the company, where more of our results, whether it's adjustments to book value up or down or adjustments to earnings, are more a function of what management is doing today rather than what our prior managements did.
I'm going to open it to the audience. I want to sneak one more question. I know you're not giving EPS guidance, but maybe I'll ask this. As you think about 2017, your own view of your earnings per share all in, has it changed much post the fourth quarter reserve charge and the ADC?
I was going to say the ADC is what changed it. As you know, we came out with some guidance on the ROE of the core portfolio at the investor day in November. There we gave guidance of 10% for the year to come. We've adjusted that to 9.5% in the earnings call yesterday. The reason for that is that we entered into the ADC with Berkshire in between and transferred the reserves to them, we lose the investment income on that. The ultimate accretion to a higher ROE comes also in the year 2018 as the state regulators and the rating agencies recognize the full amount of economic capital released by the ADC.
Sid Sankaran, the CFO, mentioned, I think, that there's a $2 billion net Release of economic capital when you look at the negative of the prior year development and the positive of the adverse development cover, but you don't get that overnight. Yeah, the adjustment to that guidance was based on that transaction, no change in our view on the fundamentals in terms of the pricing environment or our expense trajectory. However, it does reflect the higher loss picks that we did in the restatement of 2016, 2015, and are projecting in 2017. Now, obviously, we're writing a whole lot less of that business, the effect of that loss pick is less. We have put a very sizable increase in the loss pick for 2017 in that projection and that guidance.
Got it. I guess one difference between your core forecast and many people's models is we had assumed adverse development would continue. From our standpoint, that's the other adjustment we have to make, but that wasn't in your numbers at all anyway.
Correct.
Let's open it up to the audience. If you have any questions, right down front.
Hi. Can I come back to the question about segmentation?
Could you just explain in a bit more detail what was going wrong at a micro level with underwriting? When you talk about increasing segmentation, what are you talking about actually doing now that you maybe weren't doing a year or two ago?
On the segmentation, I was talking literally about how we do the bottom-up reserving. If you looked at the year-end 2015 reserving, we had about 800 different loss triangles that were driving the segmentation, and they were in many cases overlapping in terms of common risk factors like commercial auto, which would touch many of these. By re-segmenting, we reduced the number of segments to about 220. One of the benefits of that is you start to see emerging trends quicker because you reduce the noise and increase the signal if there's a statistically significant pattern there, which wouldn't necessarily show if you've got it in a more fragmented set of segments. The art is having the right number of segments, because if you look at aggregate loss triangles and have too few segments, you don't get the micro insights.
It's not that there's a right or wrong number of segments. It's just that we made a major change in segmentation in the last year, which gave us a much better read on the early emergence of adverse trends than the previous more fragmented segmentation did.
One of the things this segmentation highlighted was in your core business, your life business, the ROE there has been lower than I might have suspected. What's going on there, and can you improve that number?
Let's make a clear distinction between the life business and the retirement business.
Right.
I think that if you look at our consumer business, as we like to talk about it, we've got a very good year last year in the individual group retirement and the personal insurance, and a big turnaround in personal insurance. On the life business, it's really an interesting story of real change because a very large part of that business was effectively in runoff. It's in legacy today. The large life reinsurance transactions have freed up a lot of capital from that to redeploy elsewhere in the company and for our capital returns. What's left of the life business is really a much more modernized, more automated business. There's a fair amount of CapEx in making it a more digital experience. If you look at that, it's a business where we've shed our owned distribution and gone entirely to third party and direct.
We're now the number one direct writer of term in the U.S. We've reduced our marginal costs. We've really focused on products where we don't take a lot of interest rate risk and reduce the amount of expense in that. It's going to be emerging, I think, in the next year or two as a good contributor. It's a relatively small business compared to the past. I wouldn't overstate the scale of that in your projections. It's a business which I think we do well but I think has historically had a very large amount of capital on the in-force, which we've now largely reinsured.
Got it. What other questions do we have? Right in the middle.
Yes, thanks. You talked about the pressure you're seeing from the rating agencies. I know you're on negative from AM Best. I'm curious as to what your thoughts are on a potential for a downgrade from AM Best, and then what the impact might be relative to your business going forward should you get that downgrade.
I think that all of the three other rating agencies, S&P, Fitch, and Moody's, have affirmed the ratings that we care about most, which is of the operating companies. I think that Best will take a little bit more time to digest the huge amount of information that we have shared with the market in the last 24 hours. I'm very optimistic that they'll see the combination of improvements in underwriting mix, the power of the ADC to provide greater certainty around the historical accident years, and the improvements in operating metrics, including expense ratio. We also have over $8.5 billion of liquidity in the holding company, which AM Best in their framework needs to weave in later. If you look at S&P, they have a group-wide model which takes into account all of the parts, it's easier to solve for what they're looking for.
I think that the dialogue with AM Best is one which we have a lot of optimism over. It's critical to us because, as I stated over and over again a year ago and from the outset, our most important Promise to our clients is our claims paying ability. We view defending that rating of the operating company as a critical feature of what we do. We believe firmly that the company is better capitalized and lower risk today than it has ever been in its history. We recognize that there's so much change that we've imposed on you as investors and other stakeholders, that they need to really study the numbers to get completely confident as we are, that we've taken the proper steps to mitigate risk. No, it's very important to us.
The federal regulatory environment seems to be changing. The question I have is what would be the implications for AIG if you were no longer a SIFI? Because it's possible.
My answer to that has not changed, which is it'll be very modest. It's about number 10 on my list of things that keeps me awake at night. I think that I'm not at all ideological about regulation. We have over 200 regulators, every state and every country we work in, and in many of them, too. Our clients like the fact that we're regulated. It gives them more assurance on our claims paying ability. We have become quite good at figuring out how to work with our regulators, not against them, to give them confidence that we're doing the right thing by our clients. We think there's a certain skill in that. It creates barriers to entry to disrupt us, and you're going to hear from, I think, some fintech companies this afternoon.
Absolutely.
Good luck them dealing with 200 regulators. I think that it's an important source of competitive advantage. Federal regulation to our industry is obviously a relatively new phenomenon. It has not been as intrusive as people from the outside in thought. It was really important when the Collins Amendment, or the amendment to that amendment came along that allowed the application of Dodd-Frank to insurance company to be different than banks. Our balance sheet was very different than Met and Prue, because remember, they have almost twice the assets to equity that we have. That leverage created some concerns as well as their very large VA book. I think that for us, it was never a constraint.
The rating agencies were the key issue, and that's why we've been very focused on them and the state regulators to make sure that they understood the balance between operating company financial strength, holding company support for those operating companies, and how that all came together to having a great claims paying ability and with our tax position, a huge amount of free cash flow to return capital to shareholders while leaving behind a very strong balance sheet.
Has your view on potential for a lower tax rate changed at all?
Lower tax rate, I don't want to speculate whether it'll happen or when it will happen. We still think it is accretive to value for us. I think that on a relative basis, it's less important to us than some of our competitors, because we have a large DTA and we don't pay federal tax for many, many years. The impact of the DTA of different tax rates, I think Sid has commented on in previous discussions, so I don't want to repeat that. I think that the interesting thing is when you compare ROEs to some of our competitors that are Swiss based or Swiss domiciled, I think it'll help people to make pairwise comparisons of ROEs on an after-tax basis. We show that 9.5 number I quoted earlier is imputing an expected tax rate of 32%.
I think that as people start to do comparisons, it'll be easier if the U.S. tax rate is more normalized relative to foreign jurisdictions and probably better for the system.
Yeah, we got some questions. Let's go in the middle, then we'll come down to here.
Thanks. Just getting back to the second part of the previous question. Sounds like you're very confident as it relates to AM Best, but what would the implications be if you had a downgrade there?
Yeah, I think it would certainly impact certain lines with major accounts. I think it would be something that we have had to have contingency plans for before during the financial crisis. The big difference is that we have a very, very liquid holding company and ways to deal with it, I think that will mitigate it. It's a serious issue that we are watching, but we have no reason for concern. If you look a year ago, they also put us on watch, and they took us off watch when they fully digested what we had done at the end of 2015. I have a lot of confidence that if we continue to be very transparent with them about the risk mitigation benefits of the ADC, the modest pace of capital deployment on buybacks will be fine.
We see no problem in deploying the additional capital that the buyback authority we just announced. Given the pricing action yesterday, I'm quite anxious to deploy that capital in Q1. I don't see that as a factor in their minds so much as the operating earnings improvement that we're on target to do. I think walking them through the mix shift, how the written earns in in this current year, what it does to our interest coverage ratio, what it does to our combined ratio within the calendar year, I think is going to be the most important factor for them.
Scott, you had a question?
Well, yeah, that was my question as well. What I think we may be looking for is a little more specificity in terms of what you might see. I understand that you're looking for a positive outcome and you want to emphasize the holding company liquidity. If AM Best, if there were an adverse outcome, what will we see, do you think, in terms of attrition, % of accounts that may drop off lists? How much will premiums go down? How could you quantify that for us in a better way?
Well, I don't want to speculate about it. What I would say is that if you're looking for test cases, this company in 2009 had a lot of ratings uncertainty and a lot of existential uncertainty at that time, and yet our client retention was over 90%. Our average client relationships date back 18 years, average on 18 years. These relationships are multi-line, and they have a deep understanding of us beyond just the rating. I think while it would be negative, I think it really is about the criteria as to why it's negative, and how clients understand and look through that, and obviously the brokers. I think this company is extremely resilient and the quality of those relationships are very strong. I think it's incumbent on our team to communicate the fundamentals of this company and go past the headlines.
I think that I acknowledge that we've unloaded a lot of new information in the last 24 hours for people to digest, I'm very confident as people really digest it, they'll have the same degree of confidence that my management team and I do on the improving risk profile of the company and earnings profile of the company. I don't lose too much sleep over that, but I want to make sure that we are in close dialogue with them as the last remaining of the four rating agencies to opine.
Other questions? Yeah, Andy.
Can you comment about your logic or rationale for selling your mortgage insurance company? It seemed like it would have, at least at this point this year, would have provided a nice stable amount of earnings for you to move forward. Are there any other things that you're looking to sell, and why and why not?
That's a great business. We took it from number 5 in the market when I showed up in 2010 to number 1 in its market. We invested in upgrading the technology and the risk-based pricing framework, the distribution model. From a managerial content, we, I think, fixed this thing up to be really good. It's never been particularly operationally integrated with the rest of AIG, so it was always an option, I think, to separate. We got a compelling bid for it, especially when you couple it with the quota share reinsurance we have with it. We retain a bunch of the earnings of it, which we like, and we have other ways to take housing risk through reinsurance as well as investment in private label mortgage-backed securities on the left-hand side of our balance sheet.
We felt that we got a very good transaction done with Arch. We also looked at the outlook for the FHFA pricing, which prior to the election at least, was not as favorable because they'd reduced the pricing. The biggest competitor for United Guaranty was the FHFA.
We got time for one short question if there is one out there. We just got 30 seconds left. Let's end it here if there's no more questions. Peter, thank you very much for showing up today. Appreciate it.