Good afternoon, everyone. Thank you. I'm Jay Gelb, the U.S. Insurance Analyst at Barclays. The format of this track will be a fireside chat with David Herzog of AIG, who we're very pleased to have with us. David is AIG's Executive Vice President and Chief Financial Officer. He successfully led the financial functions of AIG during its substantial restructuring and repayment of U.S. government funds. AIG is among the world's largest property casualty insurers and also has a substantial U.S. life insurance and retirement savings business. David, thank you for joining us today.
Thank you. Thanks for having us, and thanks, everybody, for coming today.
Let's start off with the big picture on AIG. What's your perspective on where AIG is now strategically and financially, and where is it headed?
Thanks. Really from a couple of different perspectives in terms of financial strength, and then we could talk about where we're at strategically. We're starting from, today, a position of real strength. Strength of the balance sheet, the liquidity. There's still work to do on the infrastructure, but we've come an awful long way. Our operating companies, which are really at the core of the franchise, are very well capitalized, well-positioned, and what we have done is established a very good discipline to move capital around inside the company to really improve the fungibility of capital. We've got much greater insights into the resiliency of capital liquidity. We can talk about that, I'm sure, in terms of how that's come about. Our capital position is strong. Our leverage is in very good shape.
We're a little over 16% debt to total capitalization, which when we started this journey was quite a bit higher than that. We've been very successful. The team's been very successful at giving us and putting us in a position of strength and flexibility, which is good. From a strategic standpoint and an overview standpoint, Peter is coming up on his one year of being the CEO in the seat. He's taking a very thoughtful and deliberate approach, engaged in a strategy dialogue with our board. I think his shareholder letter was very telling in terms of how he's thinking about the business. Peter thinks about and has brought with him a perspective of and a balance of growth, profitability and risk. I think we're bringing those concepts to bear as we think about strategy.
We talked a little bit about sculpting of the business and certain, whether it's activities or businesses that may not be the best fit for our company going forward. Again, that's part of the dialogue that we're having with the board. I think from a strategy standpoint, we've got, again, good sound financial position from which to build on and a clear line of sight and a good process, a supportive dialogue with the board about the strategy, where we're going to compete, how we're going to compete. I think the other thing that's really important is the ongoing build-out of our data and our science team. The use of data science to help shape how we underwrite, how we think about the business, how we serve our clients is a very important and growing capability in the firm.
The function itself is relatively new to AIG, but it's an important part of how we see innovation coming to our firm. It doesn't have an exclusive on innovation, but it is certainly a catalyst for it. It partners with the business.
It's a very helpful start in terms of framing the business. Let's talk about the property casualty operation. One question I often get from investors is how AIG could further improve the underlying accident year property casualty underwriting results. It might not be at top-tier levels yet, what can AIG do to improve those?
Yeah. It is a journey that we are on. I think John Doyle, who runs our Commercial Insurance businesses talked a bit about it on the recent earnings call, that we continue to strive for sustainable improvement in the underlying profitability. It starts with the loss ratio, the accident year loss ratio, and the improvements that we're making. It's driven, John drives it really through the business mix, where we're competing, how we're competing, where we see advantage, where we see pricing opportunities. The shift in emphasis from premium volume to the value of business that we are producing and writing, and the focus on risk-adjusted profitability was an integral part of actually giving people the tools and the framework for making rational choices about how you assess whether or not you want to write a piece of business.
Historically, we would've written volume for the sake of volume. I'm not being critical. The past would never do that. It's just a statement of perspective. Now we reward people for writing profitable business. On the margin or on an incremental decision, if we can't cover attritional loss or you can't cover the losses that you believe based upon, again, a growing level of science and data, it gives people the courage to not write a piece of business. Again, that coupled with the more dynamic capital allocation, we then will redeploy the capital elsewhere in the business, or we'll get it up to the holding company, which we've been pretty effective at doing. The business mix shift is a big part of it, John's team and the science team have developed quality indices around accounts, quality indices around brokers.
In terms of the quality of business, the predictive analytics around us making better underwriting choices. Again, there's no one easy lever to pull, but at the forefront, it really is at where we face off with our clients. I think one of the other big investment areas that are paying off in a very meaningful way are around claims and claims management, and the consistency of capturing claims data across the firm. I think Peter and John have talked about how we pay something like 1 million claims a month, which equates to about $120-plus million a day. It's big money, not only is it big money, there's a lot of data that comes with that.
It's about capturing the data and capturing it once, right the first time upfront in a way that is usable and that it can be accessed across the firm. Those are tools that are new and will help drive the profitability of that business. Again, John sees, and I think there's reason to believe, it's a gradual journey. It's a one to two points a year. We were a little short of that in the first half. John sees with the other pricing environments and where we are growing and whether that's in the Financial Lines, the M&A book. Our international property book has been a big, important growth engine for us. It tends to have maybe not as much cat pricing sensitivity like you see here in the U.S. We've reoriented how we run many of our lines of business.
For example, property, that's run as a global book, as opposed to a collection of books that were written across various countries. That has enabled us to invest in the kind of engineering centers and the engineering expertise that not only helps us make better underwriting decisions, but also helps our clients mitigate loss to the extent we can. Those are some of the big tools and big levers that John and the team are pulling.
Thanks, David. AIG's property casualty reserves have had to deal with some consistent, although modest, reserve strengthening. Why is that, and when do you think that trend could change for the better?
Yeah, our reserves have been a point of a great deal of focus outside the company and inside the company as well. We do make our best estimate based upon all the available data. We look at all of the long-tail complex lines of business. There's about $45 billion or so of reserves, something in that order of magnitude, and we look at every one of these lines of business in detail in, we call them deep dives, detailed valuation reviews, and we go deep. We avail ourselves of all sorts of subject matter experts inside and outside of the company. As data emerges, we look at it in as wholesome a way as one could possibly look at it. We do have a thorough vetting. We make our best calls.
In this last quarter, I think we were disappointed with the emergence in our commercial trucking business. The team have taken a hard look at it. In some of these areas, we tend to participate in excess layers, the emergence may not be as fast or quickly as in some of the more lower tranches of risk towers. Again, we took a hard look, took a step back, looked at the why and why now, to satisfy ourselves that we're not over or underreacting to it. I think the philosophy is to get it right, get it down the middle, have pluses and minuses. Even with the net adverse developments.
There's still pluses and minuses that have taken place. We've had some adverse, particularly in some of the older accident years, that we've dealt with and we'll continue to do so. We try to get it right so that we can take some of that noise out of the system.
The reserve mix appears to be shifting as those older accident years get paid off. The newer accident years become a bigger portion of the reserve base. Should that help address the issue mathematically?
Yes, it does, and it should. Part of that, even some of the older accident years, for example, in the workers' compensation book. This is a line of business that is complex. It's highly regulated, and it is competitive. It's a vital part of the coverage that our clients need. It's an area where science, technology, data lend itself to help better inform not just actuarial judgments, but also have helped bridge not just reserves, but helped bridge claims mitigation, coming up with claims settlement strategies that have been, A, developed, B, been able to have been quantified and pushed down to people that are accountable for specific results and specific quarters so that we can track progress against initiatives that, again, years ago didn't exist. They not only exist, but they're measured and tracked, and we can monitor progress against those.
It's the use of data and the sciences in bridging across longevity risk. A good example, we have longevity risk in several parts of our product portfolios, whether that's in workers' compensation, whether that's in terminal funding and the likes, or in some of the pension buyout business. You have those same risks. We have subject matter experts. With the way the company's being run, we make much better use of internal subject matter experts. We'll continue to improve on it and continue to leverage the learnings that we're getting out of what started with a desire to make better, more informed reserve judgments, actually now helping guide how we actually run the business and risk selection and claims management.
A near-term topic that we've all been hearing about is the port explosion in Tianjin, China, on August 12th. Outside observers have sized this as potentially a $3 billion loss for the insurance industry, or probably larger. Any perspective on what AIG's exposure might be at this early stage?
We do have commercial exposure in China, so we do have both companies inside China and our multinational book. We don't comment on specific losses. We'll cover to the extent it's appropriate. We'll cover that off in the third quarter. We do have exposure there. We do have clients that have losses there.
Directionally, do you think it's within the realm of a typical large loss?
Yeah. It's hard for me to comment on it till I see more data and see again, as we run through the process, some of those areas where, like what happens in many large complex claims, data is maybe not as quick to come to the fore. We'll be thorough as we think through, A, what our exposure is and, B, how we think about how we reserve for that.
Let's turn to AIG's U.S. life and retirement savings business. The first thing I'd like to focus on is how is AIG addressing the impact of sustained low interest rates in the life and retirement savings operation?
Obviously it's a challenging environment for long-term savers in our economy. It is a challenge. That said, we have been and remain very disciplined on new business that we write. We have a developing but reasonably well-developed framework for looking at profitability of new business. We look at the value of new business. We look at the cost of capital associated with various product lines. We're very focused on maintaining our margins on new business. We've filed products with much lower guarantees, and we've repriced. For example, in our fixed annuity book, it reprices literally weekly, and it's done so with an eye towards maintaining a return. We've been very disciplined about not overemphasizing, not growing that business. Not so much that we couldn't earn our cost of capital on that business today. We can. We are.
Rather with this level of nominal interest rates, we've been cautious about not overbuilding a book of business that would exacerbate a disintermediation risk for the firm. There will come a point in time when the macro demand for that product, fixed annuities, will become much more robust. A 10-year rate approaching 3%, we have seen historically that spurs demand, macro demand. It really is around pricing discipline and also pricing discipline on reset when the annuities come up for their window renewals and we reset. We've been very disciplined about moving interest rates to the extent we can. You move them, in many cases, down to their minimum. I think it's roughly about 70% or so of that book is at its minimum guarantee.
We're still making money, we're still making a spread on it, we've been disciplined about bringing down the rates. Again, it gets back to a broader product mix. Our variable annuity business continues to be very attractive. We have a broad distribution platform and the products. We were not major players in the variable annuity space during the early days and maybe where some of the benefits may not have been priced exactly right. Our back book is really fairly modest. The product features, the risk management that goes into the product design, the best risk management in variable annuity is at the design table, not once you have it on the books. The team's been very careful about growing. Listen, people are worried about outliving their income, and that product set serves a very specific and important need for our clients.
Having that product and having good quality distribution is an important part of that. We've been, again, going at it in a very balanced way.
Thank you for that. Well, let's turn to capital management and dividends. This is a real important focus area for investors. AIG has been a substantial repurchaser of its shares and recently doubled the dividend. How should we think about the pace of buybacks going forward, especially taking into account the proceeds from the sale of the aircraft leasing business, ILFC, to AerCap?
Yeah. It is one of the most important levers, not the most, but is one of the most important levers we have for creating value, is how we deploy capital. We have a framework that we engage with our board and a dialogue we have with our board about how and where and what quantum to deploy capital. We always look for organic opportunities, we generate sufficient capital. We'll look from time to time at inorganic opportunities that enhance a capability, enhance a skill, bring us something that is easier bought than built. Generally speaking, inorganic or acquisitions that are not capital intensive. They're not necessarily transformative. We've purchased a number of, for us, relatively small companies. They've been very important acquisitions for the respective local businesses.
The Ireland acquisition we did was very important for that business, but it wasn't transformative for AIG, but terrific for that particular country. Then we think about capital deployment and share buyback. The sources of capital come from some pretty well-defined, well-established sources inside the firm. The dividends distributions up from the operating companies, the monetization of our deferred tax asset, which comes to being in respect of the tax sharing payments that are made from the operating companies up to the holding company. When we file our tax returns, we're utilizing net operating loss carryforwards or foreign tax credits. That's a real tangible source of cash that comes up from the operating companies. The monetization of the non-core assets, which is what you referred to. The capital planning process, we have a plan. We have a policy.
We have a plan. We review it with our board. The Fed, they review our process, but they don't review or approve our plan. As we generate deployable capital, we then again enter that dialogue and either incorporate what we've done into the then current plan, or in this case, with the $5 billion of additional authorization that we sought and was approved by our board, it was on the basis of we had had a material monetization of the AerCap shares, a material monetization of a portion of our PICC shares, which that's been a terrific relationship, but it had appreciated to a great extent, and we wanted to take some money off the table and diversify a bit. It's still a wonderful relationship, and we value it greatly.
We had some other Springleaf, which was a sale of remnants of one of the divestitures we had done back in the restructuring. Those are things that had occurred, but none of those had been incorporated into the plan, the original plan. There's a methodology to that because what you don't want to do is plan on a monetization, incorporate your capital actions, you build a plan about it that's based on contingent funding. That is not a formula for success by the Fed or frankly, by our board. We're much more prudent about that. We will continue to deploy those authorizations. You have seen our past practice of getting an authorization and then in fairly short order deploying it.
I would expect, I think Peter commented on the call that we'll either have it utilized by the end of the year or shortly thereafter, I think that's an appropriate framework for how we think about it. We'll update the world.
How much is left in the current-
Well, as of June 30, which is really all I could best comment on at this point, of the entire $11 billion that we had authorized for the year, we had $6.3 billion remaining as of, I think it was as of June 30, or maybe that was as of the time we released earnings. I think it was as of the time we released earnings in early August. $6.3. It'll be that $6.3 that we utilize between here and either the end of the year or shortly thereafter. Maybe it spills over into the early part of 2016. It's all facts and circumstances based. It's a dynamic capital model that we use to decide how quickly, how, excuse me, when. We are deploying. We will deploy it, and that's why we have it.
That's why we sought to generate the capital and then get it deployed.
I think investors very much welcome that, especially with the stock trading at 80% of book value.
Well, we're aligned. We see it in a similar way. When it's trading below one's view of value, whether that's view of book value, whether it's adjusted book value, whether it's a range of intrinsic value, when it's trading at a discount to that or an attractive level to that, it becomes a compelling use of capital.
Okay. AIG has been designated a non-bank SIFI, which means it's regulated by the Federal Reserve. What are your expectations for the non-bank SIFI rules and potential impact on AIG?
It's interesting. Our relationship with the Fed has been, first and foremost, been constructive, both on our side with the Fed and I think with every interaction I've ever had with varying parts of the Fed, whether it's the New York Fed or the people in the Washington Fed. It's constructive. Today they are, I would characterize them as our supervisor because there aren't rules. They'll regulate us when there's a set of capital rules and supervisory rules that have been issued. They're in the process of developing capital standards. We, like other market participants that are similarly designated, are working to try to educate the Fed, come up with a sensible standard. The amendment to the Collins Amendment made it possible that the Fed can differentiate banks and insurance companies. I am encouraged.
I'm heartened that my conversations with the Fed, they acknowledge that banks and insurance companies are different. Their balance sheets and the balance sheet velocity is different, therefore the capital framework and rules should likewise be different. They have a challenge because they have to come up with a standard that is appropriate and manageable by companies that do not file regularly U.S. GAAP financial statements. Today, when the banks are regulated, they're regulated on a U.S. GAAP basis. The new challenge is that the Fed's got to come up with a capital standard that's adaptable by not only those of us that are U.S. GAAP filers, but those insurance bank holding company or thrift holding company institutions are not U.S. GAAP filers.
We're coming up, we're helping give constructive feedback on what kind of standard that might be, what might that look like, how might that work. In the meantime, Jay, we're operating a company really according to our own true north. We run stress tests. We update those stress tests. It's on a U.S. GAAP basis of solvency. We look at solvency. We look at liquidity. We look for the binding constraint in the system, so to speak, because that's one of the lessons learned from 2008 was that you can have plenty of capital maybe on a consolidated basis, but if you don't have the capital in the right place at the right time, in the right economic circumstance, you could find yourself in a bit of a problem. We have our own true north.
We continue to develop that true north, building out the infrastructure, the robustness of the test, because what we have learned, what I have learned is that it goes beyond just the numbers. It's not just a quantitative test. It's very important. I think the Fed sees it as what's our process, what's our control, what's the level of dialogue and challenge and review both internally and with our board. It's a learning process. I'm glad we started early and have continued to build out that discipline and build out the infrastructure because it's important. We want to be ready. Whatever the standard is and whenever it becomes applicable for them to become our regulator, we want to be ready.
Any sense of the timing on that, either for the rules or the initial stress tests for the non-bank SIFIs?
Not that I would place any real weight on. As best I understand it, there's going to be a process. The rules will be drafted, they'll be released, we'll have a chance to comment. They'll review our comments and those of, again, other market participants. I would expect there to be some type of quantitative impact study or field test so that they understand the possible implications of what the rules might be. Once the rules are finalized, it's then the following year that they become applicable. There's a process to go through. I wouldn't expect the rules to be finalized this year. Does that then translate into a 2017 at the earliest? It's conjecture on my part, just I'm telling you what, from the perch I sit in, that's kind of how it looks.
2017 for a stress test.
Yeah, I think so. In the meantime, we continue to run our own true north, and we go through quite a bit of rigor. Our board puts forth very high standards on us, and we have that dialogue with them. And frankly, on ourselves, our own true north. We take a lot of pride in the resiliency of the firm, and want to make sure we're ready. Again, the whole idea is to run the tests and operate the company such that when the rules are finally formalized and applicable, there's not some sea change in how we're running the company, so that it's as seamless as possible. There's no question there's going to be a lot of work to do.
The whole idea is that the fundamental operating parameters of the firm are as best we can have them at the time the rules become effective.
Thank you for that. The next topic I want to focus in on is return on equity. AIG's operating return on equity in the first half of 2015, excluding AOCI, which is essentially unrealized fixed income gains, and the deferred tax asset was 9.3%, or if we look at it on a normalized basis, it was 6.7%. This result is below many property, casualty, and life insurance peers. If you look out over the next few years, what do you think a reasonable expectation should be for AIG's return on equity?
Yeah, we tried to come up with a normalization of our ROE. We have an operating income definition that's been very consistent. There are some items of variability or volatility that will occur from quarter to quarter. We tried to put out a "normalize" for that. As a benchmark or a baseline for us to measure our progress on the fundamentals. The baseline, which was 7.4, and I think the first half of the year was somewhere around 7.3, I think the first half, is as much for us to measure our progress against ourselves, not so much as a measure of what that number is against somebody else's non-normalized number. We also, just as a point of fact, we had done quite a bit of capital gains harvesting in the past because we had some very valuable tax attributes, capital loss carryforwards.
In certain targeted books of business, we harvested capital gains, which is good for all of us that are shareholders because we captured quite a bit of economic value in doing so. Now, the consequence of that is that our U.S. GAAP reported operating income is hurt because of the reinvestment in lower rates. There was about, I don't know, 70, 80 basis point effect on headwind on it. We did not adjust for that. That's just a simple fact. We've laid out a goal objective of improving ROE 50 basis points a year for the next three years off that baseline. Again, as a measure of improvement of the underlying fundamentals. Again, the key levers to that, obviously, there's a lot of leverage in improving the underlying profitability of our Commercial Insurance business.
Essentially, one point of combined ratio improvement equates to about $200 million a year in pre-tax earnings. There's the loss ratio improvement and expenses, which is a nice segue to why the second goal and objective that we've laid out is an improvement in our general operating expense levels across the firm. We've laid out a 3%-5% net improvement in operating expenses each of the next, including 2015, each of the 2015, 2016, and 2017. There are a number of, I would call it a program of work that is well underway across the firm. It's not going to be 3%-5%, I call it mayonnaised evenly across the firm, but it'll be done in a very targeted way.
That's a big lever because our operating expenses total about for 2014 was just shy of $12 billion worth of operating expenses across the whole firm. Again, that's a big lever. Like I say, there's a broad commitment to that. We're all tied together. Our incentive programs and our performance, my performance, Peter's performance, the entire team's performance will be judged on how well we manage that. Again, it's not going to be a straight line. It's not like clipping coupons, but it's a body of work. We announced recently we've made some important changes to some of the benefit programs of the firm. We didn't quantify that, but obviously, that was a meaningful step, and that's just one of many, many areas where there's plenty of opportunity to improve the efficiency of the firm.
Right. 50 basis point target improvement in return on equity for each of the next three years with a starting point of where?
We said the 7.4 based on the normalizations that we had done. What's interesting, we had the timing of AerCap, for example. It was a high ROE asset, nonetheless, I would just say no regrets on having sold it. It was a non-core asset, I think the team clearly made the right choice. It's off the 7.4, and we'll go from there. We'll have headwinds from time to time, AerCap, in terms of an ROE progression, is clearly a headwind of 30-odd basis points. That's our goal.
Okay. 2017, hopefully targeting around 9% normalized. To me, that's pretty close to 10. That's putting you in a pretty good spot, especially with the price-to-book multiple and the stock currently.
It's as much about managing or trying to influence the cost of capital as it is the absolute return on capital. We've done an awful lot to appropriately de-risk the firm. We have to take risks to make money. It is about managing both the cost of risk and managing our return on the capital that we're deployed.
I'm afraid we're essentially out of time. Were there any final thoughts you'd like to leave the group with?
No, thanks. I think you've hit on the right topics. Thank you.
Appreciate it. Please join me in thanking Dave Herzog from AIG. Well done. Thanks so much.
Thank you.
Really a pleasure.