Hello. Thank you all for coming. We'll get started here. I'm Joshua Shanker. I cover the property and casualty insurance group in the U.S. for Deutsche Bank, and we're really pleased to have Peter Hancock, CEO of AIG, presenting at this time. Peter has come from the banking world to the insurance world over a very interesting path. He started at J.P. Morgan, where he was head of derivatives, head of credit and fixed income, and migrated to being the Vice Chair of KeyBank. Of course, his services were really needed by AIG post-financial crisis, when the financial products book and the derivatives book needed to be unwind to take the risk out of this company. Peter joined the team at that point in time.
As that moderated, and his skills were obvious, he led to running the P&C business at AIG for a number of years and ascended to the role of CEO, where he is today. It's a different company than it was five years ago when he joined. Clearly, he's put his imprint upon it. Peter's going to give some prepared remarks, and then I'll ask a few questions in this sort of this whole fireside chat setting. I hope that you, the audience, have some questions for Peter, and he'd be very happy to answer them. Please welcome Peter Hancock to the stage, and thank you very much.
Josh, thank you very much, and it's a pleasure to have a chance to speak to you about AIG and also hear from you any questions that I can answer. It's been an extraordinary experience to be a part of the team that has led to the reshaping of this company. It's been through a period of transition like no large company that I'm aware of in terms of the degree to which we have focused it on its core competence and exited distractions, non-core activities that either distracted management's attention or, more importantly, diverted capital away from what we do most productively. If you look at the scale of that, it was taking our balance sheet from a highly levered balance sheet pre-crisis to one of the most conservative balance sheets in large cap financial services today.
Derivative notionals of over two and a half trillion pre-crisis to about $130 billion today. Compare that to the bank SIFIs, they're between $20 trillion and $60 trillion, with a T. We've effectively done by the end of 2012 what GE announced they will do sometime in the future. Give you a sense of the de-levering of the company and the refocusing of the company during this transition period. We're now shifting to a focus on sustainable growth and making sure that each of the components of AIG, which today is still by many measures the world's largest insurer, operating in over 100 countries and jurisdictions with about 90 million customers. To make sure that all of those pieces fit together in a way that best serves our clients. That's why we have organized our company in a way that very much focuses around client needs.
We have a commercial division and a consumer division with three regions. That matrix effectively ensures that we take an outside-in view for our clients as to what it's like to do business with us. To do that, we have to make sure that all of the pieces fit together. We look skeptically at whether each of the activities we're in are reinforcing each other, whether they provide diversification of capabilities and diversification of capital utilization, whether there's shared technology that makes us stronger together. We're agnostic as to what pieces belong and which pieces do not. In our shareholder letter, I made it very clear that we look at that. We start from the perspective, how can we most effectively be our clients' most valued insurer?
To do that, we need to continually invest in innovations so that we capture the benefits that data can give us to better underwrite risk, better manage claims, and better service our clients, and do that cost effectively. There has been a tremendous amount of innovation going on in financial services in general, but this company and the insurance industry in general is a very data-rich environment. To give you a sense of the scale of that, we pay roughly 1 million claims per month. We pay roughly 134 million claims per day, whether it's a $50 claim for a lost suitcase or $1 billion for a microchip plant that explodes. Between those bookends, there's a huge amount of continuous learning, which makes us smarter underwriters.
The way we synthesize that information and get better at what we do and share those insights with our clients is our value added. It's not simply about risk transfer. It's about helping our clients manage their risk, manage their uncertainty, and empower them to grow their businesses, or as individuals, to plan their retirements in a way that uses our insights about what they should fear, what they should not fear, what they should insure, what they should not insure, and what is the most cost-effective products that can help them do that. We need to deliver that through channels that are most convenient for them. The good news is that we have a very broad range of distribution channels that we interact with our clients with. We're not beholden to any one distribution channel. They differ obviously by client segment, by geography.
Here again, there's a tremendous amount of change going on. As all of you as consumers know, we expect a level of digital experience that was unheard of even five years ago. We need to invest to give our clients and our employees the digital tools to really improve the client experience in doing business with us and understanding the risks that they have. We're investing in data, we're investing in science, we're investing in digital for our clients and our employees, while at the same time acknowledging that our financial performance needs to improve. We made very clear when we talked about our financial goals, that while we think we've made tremendous progress, we can do better. I want to acknowledge that. We set three important financial goals for it.
The first was to grow book value per share in each of the next 3 years, this year, the year after, and the year after that, by more than 10%. That's inclusive of dividends. We have a very, very small dividend today. Our dividend yield's less than 1%. Should we increase our dividend, we would adjust that book value per share growth rate to accommodate the cumulative return, combining book value growth and dividends. Just to give you a sense of how we think about that target. We want to do the right thing in terms of dividend policy. The second goal we've set is that we intend to reduce net general operating expenses by 3%-5% per annum. That's net of investments. As I mentioned, we've got some very exciting investments in our organic growth opportunities.
We also have a number of savings opportunities that have been in flight for some time and new savings opportunities as we think about how we can better run the company today going forward. 3%-5% net expenses, which includes accommodation of our investment plans, to continue to grow our business and modernize our business where it needs to be modernized. The good news is last year, we made a good head start on that with a GOE down 3.5% on the year before. It's not a standing start. We've already got that GOE improvement baked in. The third goal is the return on equity improvement that we've set out as a goal, which is 50 basis points annual improvement in our ROE. We announced in prior disclosures that we normalized last year's return on equity at about 7.4%.
There are a couple of headwinds and a couple of tailwinds, which together netted out to about 100 basis points overstatement of the normalized rate. You may say, "Well, that's a pretty miserable ROE compared to peers." Peer comparisons are somewhat tricky with our numbers. We have a few idiosyncratic aspects in the calculation of ROE that you should all be aware of. If you're not, I'm sure Josh can lead you through the math. We've got a very large DTA, deferred tax asset, which you need to think about. We've got AOCI on our unrealized gains in our bonds, which has changed dramatically over the last 5 years as a result of capital gains harvesting.
We have sold a lot of high coupon bonds, both in the life and the property casualty business, for all sorts of good reasons, and replaced them with lower coupon bonds, which has the effect of increasing book equity and reducing net investment income. As you compare us to peers, for instance, which I'm sure many of you do when you think about our ROE and price to book multiple, it's worth just taking a look at those cumulative capital gains and what effect it's had on the portfolio yield, because we have a longer duration liability stream in our P&C business than many of our peers that you may be thinking of that have short-tail business. Just wanted to caution you about that as you look at peer benchmarking as opposed to period-to-period benchmarking.
Our 7.4 number is to help you do period-to-period benchmarking for AIG as opposed to peer-to-peer benchmarking against other companies. Just want to clarify that. We think a 50 basis point annual improvement on a capital base of over $100 billion is what we think is a prudent, sensible pace at which we can deliver results that get the correct balance between current profitability and a sense of urgency. Investing for the future. Recognizing this is a company founded 95 years ago with a long, proud history, client relationships all over the world that go back many decades, and that we're investing to be around for at least another 95 years.
Balancing short-term priorities and long-term priorities is baked into those goals, and we think we're getting this balance between profitability today, long-term growth, and the other third dimension of our framework for thinking about intrinsic value, which is risk. If 2008 taught us anything, it's that financial institutions need to have all three drivers of value in mind in good times and bad. We think about that all the time at the macro level for enterprise risk, but at the micro level, when we decide whether this product or that product, this region or this customer segment offer an appropriate balance between these three. Have we saturated the growth opportunity here, and should we optimizing for profitability?
Is this a sector which has got tremendous growth side and we should be less worried about the current profitability and more worried about whether we can build a technology platform that can scale for the growth opportunity? This sort of strategic thinking is what we are trying to embed at all levels of the organization with a governance structure that gives leaders at the appropriate level in the company the autonomy to make the trade-offs between profitability, growth, and risk that adds up to these aggregate financial outcomes that we are promising to you as our shareholders so that we can deliver sustainable growth and value of the company. We think about intrinsic value as our goal. It's informed by market value, but obviously, market value can over or undershoot based on perceptions and technical factors.
For us, the intrinsic value provides a true north that ensures that management is keeping their eye on the proper balance between short, medium, and long-term performance goals. How do we then deliver the sustainable returns that I hope you want to have as holders of our shares? In terms of short-term priorities that we have right now, we're clearly very focused on simplifying decision-making in the company, and that's a big driver of the expense initiative. We're very focused on improving the technology experience, the end-user experience for both employees and clients of the company. We're very focused on improving the metrics that we have for value-based management. Here, I would say that we've made a lot of progress, but we still have a ways to go. I am a big believer in really looking at the company through an economic lens.
Discounting both assets and liabilities at current market discount rates, really understanding embedded optionality in policies that have been sold, and to do so in an industrial-strength way where you get a feedback loop to managers that they know that regardless of what the GAAP results show, positive or negative, that we are truly adding economic value in what we're doing, both at the tactical and at the strategic level. Building the infrastructure to do that efficiently and inform people in the MIS perspective, what's good business and what's not so good business is a really important priority this year so that we're all working from the same information set. The final priority this year is to really embed this concept of being customer-centric.
That really gets to the recognition that for many, many years, AIG was a lot of profit centers focused on their own product specialization, often with competing brands and competing client strategies. Many of our clients deal with us in many products, in many geographies. Being able to integrate that and give them a customer experience where they're dealing with one AIG is something that they have indicated will really differentiate us from our competitors. I had one risk officer of one of the top 10 banks who renewed their global property program with us in 110 countries. This risk manager had been doing business with AIG for over 20 years. He said this renewal was the first time in 20 years that he felt he's dealing with one company. That was a very gratifying feedback that we're making some impact here.
I wish all the risk managers of our clients would say that, so we have work to do to get there. I would say this idea of becoming our client's most valued insurer is something that we can show at different customer segments. Let's just pick home base since we're here in the U.S., the Fortune 500. We do business with more than 99% of the Fortune 500. Capturing more clients is not the route to growth. The proportion of those Fortune 500 clients that would designate us their most valued insurer, while material and probably higher than any other insurer, has plenty of room for improvement. I see tremendous growth in more deeply penetrating the existing client base with more value-added products, but importantly, services. That's one of the areas we're investing.
Services that make us as valuable to our clients for our risk expertise as our financial capacity. What we're delivering is not just simply our balance sheet, but our risk expertise. With that, I'd like to turn over to Q&A.
Thank you, Peter. You've left me wondering who the one to four companies in the Fortune 500 are in business, I'll work on that. In your letter to shareholders, this is your first letter to shareholders you've written, you lay out, I think, your vision for the company for the first time, I think I would encourage anyone to read it. Of course, for me, one of the key tenets is the 3%-5% expense reduction. Back in 2011, the company said reducing expenses was a very important part of the long-term strategy. We spent about three or four years increasing expenses to bring the company to a position where I think it's clear now we are over that hump. To what extent is the 3%-5% number an aspirational number?
To what extent you have confidence around it, given what you know? I've made the joke that many of the people I know at AIG already feel the belt is quite tight. How much tightening is there to come, or how much of this is certain opportunities, like the cessation of the integration in Japan, the global expansion of fraud detection, new technologies for capital analysis and whatnot? How much can we expect that the 3%-5% is easy versus the 3%-5% is hard to do?
The news on this is it's never easy. I think that cost cutting is intellectually easy but emotionally hard. Growth is emotionally easy but intellectually very hard if you want to find profitable growth. Just owning that issue is important to get people to the table. I can say that the budgeting process in November and December of last year, in the five budgeting cycles I've been in at AIG, was the first time where the predominant topic of conversation was how we allocate our expense budget. Which is the element of what we control most as opposed to top line and other factors that are obviously subject to market forces. Getting the focus of the budget process on the things we control most was the first shift in mindset. Making sure that we don't make it across the board.
I like to tell management that we are paid to deliver mixed messages, which means that we have to grow in some places and shrink in others. As you go down the cascading levels of management, you'll see that we're growing in some places and cutting a whole lot more than 3%-5% in others, so that we reshape, we sculpt the company around activities where we add the most value. It's not going to be easy, we wouldn't have gone out with a public target like that if we didn't think we could do it.
To what extent do you feel that you have some tailwind on it from the expiration of certain large investments that will cease to be part of the scale? You obviously don't want to give out numbers, but do we expect that in that next 3 years, there are some step function downs in terms of expenses that long-term budgeting makes it easier to get to those numbers?
We've had one very big project that's been going on for a long time, really since the 2005 restatement, which is modernization of the basic plumbing of the finance area, the general ledgers around the world, and that's tailing off already and has been tailing off. What's still underway but will tail off in the latter part of next year is the integration of two very large companies in Japan, AIU and Fuji Fire Marine. These two companies are each over 4,000 people, quite obsolete systems, converging on a common platform, rationalizing a branch network that goes up and down Japan with hundreds of branches into one that's more effective. Under Japanese regulatory constraints, you can't jump the gun. You've got to wait till they merge before you eliminate a number of the overlapping functions. That will be a bit of a tailwind.
In addition to that, we've been migrating a sizable number of jobs from diffused locations into shared service centers, both in the U.S. and overseas. The pace of that migration is, I think, peaking this year. It'll continue, but it's about 2,000 jobs per year. It's about 1,600 the year before and about 1,000 the year before that. What happens when you've got this major migration is that you have a mirroring of job functions, and that sort of leads to elevated expenses. As you start to get over the hump, that changes. The first thing is this pooling in the shared service centers. You get labor arbitrage to some extent. More importantly, you start to get critical mass, and you start to get more specialized management of workflow, so you get workflow optimization. Then the final step, you get automation.
That's where the long-term sustainable cost savings come from. It's a multi-step process. It takes several years. Many of these initiatives were started four years ago, and so they're coming to fruition. Others have been initiated more recently and are still in train.
Makes sense. Tell me about capital generation and capital return. I think it's been almost a year that the first time I heard you talk about the subsidiaries generating about $4 billion-$5 billion in dividends to shareholders and the intention that you'll get $1.5 billion-$2 billion of tax-sharing savings as well.
In easy math, that comes to $5.5 billion-$7 billion available from standard operations. This year you announced that you intended to return $6 billion-$7 billion to shareholders in the form of dividends and share repurchases. If you've read my work, I think the numbers are higher than $5.5 billion-$7 billion. Notwithstanding that, you sold about a billion and a half of non-core assets in the first quarter of the year. It was announced two days ago that you're going to reduce ILFC's stake in AerCap. That's another $2.5 billion. There will be a $2 billion one-time capital release out of the direct investment book in this quarter. It may have already occurred at this point. That adds, to my reckoning, another $6 billion that may not have been the accounting when you thought about $6 billion-$7 billion.
How often does the board and you revisit the capital plan? How should we think about that incremental $6 billion related to the capital generation elsewhere, and That's basically it.
We have initiated a much more dynamic capital planning program.
It happens about once a quarter, and if necessary, on an interim basis between that. It's really a continual process as events change, as assets are sold, as opportunities to grow and deploy that capital for organic growth arise, as acquisitions are evaluated. Also as organic growth changes. There's significant cyclicality, and there's some pretty capital-intensive businesses which have become less attractive, and we'll pull capital out of them and redeploy that capital elsewhere, where it gets a better return. All of that lends itself to a more dynamic capital planning process than perhaps people had expected. When we look at the hierarchy of how we deploy that capital, we believe that when our shares are trading below our estimate of intrinsic value, that's a very high priority for where we should spend it.
We have also looked at debt reduction last year as an important priority, in particular the high coupon illiquid debt. We reduced debt by over $10 billion last year. That debt reduction and debt rationalization is, for the most part, over. It's really about the equity, and we're watching how it performs relative to intrinsic. I think that we will be dynamically adjusting that as markets change. We recognize the importance of husbanding capital, your capital, our shareholders in the room, very much with a view to making sure that we return it to you if it does not get an attractive return on risk. We don't want it to be idle capital. We want to return it to you promptly.
Linked to that, we are judging at which point we should return to a more normal dividend, because we have a nominal dividend today, and we see no reason why at some point we shouldn't be in the pack in terms of dividend yield, dividend payout ratio, and so on. I think we have to make that trade-off based on the economics, but also the signaling that it suggests in terms of our confidence of our sustainable earnings.
You've said explicitly that the $6 to $7 billion capital return plan had no bearing on your decisions on what to do with AerCap for the remainder of the year.
Correct. That excludes non-core asset disposal.
Would that go for the other non-core asset sales and the DIB capital release as well? They were not contemplated when the $6 billion to $7 billion was-
I think that while the sum of the DIB and the PICC sale and the Springleaf sales come to about $1 billion, I have to say that the precision with which we do that is not that precise. I think that we like to think that we run a pretty tight ship here, but that's a sort of swing number that I wouldn't say was in or out. It was in the back of people's minds before we set the plan. It got executed while the plan was being approved. I think it's very hard to-
These Excel spreadsheets are complicated things. I'll ask one more question, then I'll open up to the floor. Right now, you and a couple other companies are non-bank SIFIs here in the U.S. There are some people who believe that's a handicap. There's a theory or thought that that's the best situation for AIG right now. Even beyond that, there's the G-SII, which I mean, it's far beyond my understanding what that's going to entail at this point in time. What are the costs for AIG to get out from under the regulation of these things? What are the benefits AIG is currently enjoying because it's in the regulation pool? I realize, I think you said in the last conference call that we'll always consider all options. Right now that's not an option we're thinking about.
How does that planning go on for that long term sort of idea?
Well, I think the first thing to recognize is that there's an enormous difference in this area between banks and insurance companies. There always has been, but it's never been more stark than today. Banks have fundamentally been much less regulated than insurance companies. I've worked in both sectors. We have roughly 200 regulators. Every U.S. state, every country we operate in, and in many countries, two regulators. The system of regulation for financial services in general has obviously been severely tested by the financial crisis, and there's been many reforms to try and improve it. The creation of the SIFI designation, the creation of the G-SII designation, the Global Systemically Important Insurer, is another approach to modernize this.
The individual regulators of each of our jurisdictions, state by state, New York State in particular, but the FSA being converted into the PRA and the FCA in the U.K., every one of these 200 regulators has been looking at what worked and what did not work in 2008. There's a tremendous amount of change. Change in awareness of vulnerabilities, change in what institutions should do to be more transparent, and how much leverage makes sense. Now, the good news is that compared to banks, insurers are much less leveraged. We have less than five to one asset to equity ratio. Most SIFI banks, it's 12 to 15 to one. We have, as I mentioned earlier, about $140 billion of derivative notionals to SIFI banks, $20 trillion to $60 trillion. Liquidity. We have long-dated liabilities and more than 85% of our assets are tradable securities. Banks, the reverse.
They have short-dated liabilities, deposits, and they have illiquid loans on the asset side. Whether it's leverage or liquidity, there's a world of difference between banks and insurers, and in particular, this insurance SIFI and the bank SIFIs. Where is the complaint coming about being a SIFI most? It's on the bank side because they're being forced to de-lever. We did that before 2012. I think that we view the leverage constraint as not a big deal because we have our own true north of what level of risk makes sense. That's our binding constraint much more than what we think the future rules of SIFI and G-SII may be, because those rules have not been written yet. It's really jumping the gun to say that it's going to be bad or good. So far it's been fine.
I think that the way in which the individual regulators operate, both at the state level and internationally, is also something we watch very closely, and how having a systemic regulator helps coordinate that. That provides, I think, a great deal of comfort, especially to foreign regulators, that they're not going to be in a Lehman-like situation where capital is transferred from overseas subsidiaries overnight because there's an enterprise-wide oversight approach that makes sure that the whole company has integrity in its stress testing and so on. On balance, I think it's very important in this business to have the confidence of your policyholders. We make large long-term promises as part of our core business. The rating agencies who keep a watch on us as well.
I don't think there's anything like as much of a division of interest between creditors and shareholders in this industry than in perhaps some others. The importance of maintaining a balance sheet with a high degree of reliability through the cycles, whether it be credit cycles or insurance cycles, is very important to our clients.
Well, I appreciate the answers. Why don't we open up to the floor? I can always ask more questions. Well, right here.
Hello. Just given the history since 2008, do you find any government interference apart from the usual regulatory agencies in your day-to-day affairs at AIG? I think sometimes that may be an overhang on the stock.
I don't think that there's any interference that I think that other companies don't feel in most industries. When I speak to CEOs in other industry sectors, they all have a lot of engagement with the government. I think the word interference is suggesting that they are getting in the way of doing what you want to do. That's almost an ideological position that many people take as opposed to a pragmatic one. I'm not an ideologue. I look at whether government is partnering with industry to try and get good outcomes for our clients. For the most part, I think that's true. There are a few exceptions where you have politicized actors, for the most part, I would say where we encounter government, whether it's in Japan, the U.K., at the federal level in the U.S., or in the state level, we see an alignment.
The benefits of having a collaborative relationship with all of the government agencies you operate with, in my view, far outweigh scoring points as to whether more or less government is good. We are in the business of serving our clients and rewarding our shareholders with great returns. Getting into an ideological battle of whether more or less government is a good thing doesn't seem like a winning proposition to us.
Do you see the mortgage business as a core holding?
Yes is the simple answer. As I said earlier, we look at all of our businesses through the lens of whether they will always be a core holding, and it's an obvious one that could be spun or sold because there are several peers that trade publicly. It invites that question, and we are always looking at good questions like that internally and from shareholders to judge what belongs together. Right now, and until we decide otherwise, yes, it is core.
You talk about AIG being around for 95 years. You hope it's going to be around another 95 years. Currently, your ROE is 7.4%. It's obvious that today it's not the same AIG as it was over 95. Could you tell us in what ways it's different? What was done? There's obviously less risk-taking, less profitability, but could you go into some of the changes?
Well, I think that, obviously, 95 years ago, it was a startup company. Even relatively recently, it was run to some extent like a startup company with a lot of big personalities at each level of the organization, quite opportunistic, and very focused on growth. The 1990s was a period of rapid growth, and the market put a very high multiple on the earnings based on growth expectations. I think in hindsight, there were a lot of lessons learned that there's a phase in any given market where growth is the right thing to prioritize. You're a first-mover advantage in this or that sector, whether it's a country where we opened insurance in many countries which had barely seen the insurance industry or product.
At some point when you become a dominant force with a large market share, marginal growth is probably not as attractive a risk-return prospect as the early part. Knowing how to set the right balance between profitability, growth, and risk is something that we've learned and are applying in our thinking. It's recognizing that we need a more nuanced set of goals, growth in one sector, efficiency in other, de-risking in another, so that the aggregate is a sustainable growth in intrinsic value. We think that while that ROE number you cite, 7.4, is a useful starting point for period-to-period improvement, as I mentioned earlier, for peer-to-peer comparisons, I don't think it tells you the full picture for a number of reasons, including the capital gains realization that I mentioned.
We acknowledge that it should improve, but we also think our cost of equity has come down dramatically and will come down even more as people realize how much we've de-risked our balance sheet. We've sold over $60 billion of equity and equity-like assets from the company since 2011. Our beta should be a lot lower than it appears to be, and I think that should drive a lower cost of equity. We're focused on what we call the risk-adjusted profit spread, or RAP spread, which is the spread between our ROE and our cost of equity, and we want to have a sustainably positive RAP spread through time, and that's how we think about our business. That means that we will not just simply pursue the highest ROE business. AerCap, which we announced the sale of, is the highest ROE business.
We just sold it. It's non-core to insurance. It's a very high beta business compared to life insurance. We use a differential cost of equity as we evaluate the RAP of each of our lines of business. This was not the way we evaluated businesses in the past. We prioritize growth, we prioritize ROE, and we're now more balanced in looking at the risk profile and other dimensions of value.
Given your background in risk, how do you think about the tail risk involved of an adverse verdict in the Greenberg lawsuit, given the kind of unique nature of AIG's relationship with the government since 2008?
We talk about that in our 10-K. I'm not a lawyer, but as I see it's a compound problem. There's a first verdict that will arrive quite soon, and that could go either way in terms of for Starr or for the government. There's the question of the damages, if any. The likelihood is a first level of appeal and a final level of appeal. You've got three outcomes that could go either way and the damage amount in each case. On top of that, the compound risk of whether the government would seek to get restitution through their indemnity from us. That is affected by the arguments that each of the judges might have along that journey as to the grounds for the case going in favor of Starr.
If the grounds are ones that would nullify the indemnity, that would make that indemnity less of a concern. As we think about those probabilities together, we come up with an assessment of whether this is a cause for concern. Whichever way, it's not going to be resolved for many years. It's not an immediate liquidity event, but from a valuation point of view, that's the framework we use. I think you have as much information in the public domain to assign your own probabilities to each of those elements to come up with as good a guess as we have.
I guess, are there any more questions? There's one. All right.
Wanted to ask about the P&C pricing environment. I think over the last few years, AIG has generally outperformed its peer group, better pricing growth. What are the key drivers for that performance, and do you see them persisting for how long?
Well, as I have tried to indicate, we have been very disciplined about getting this balance between profitability, growth, and risk. That's right down to the individual desk level, where people are budgeting in terms of marginal RAP rather than in premium. In the past, if you were trying to hit a premium target in your budget, as the softening and pricing happened, you might keep on writing business. We take away that incentive to keep on writing if the pricing environment diminishes. We'll reallocate that capital elsewhere in the business. To give you an example, we have been writing a lot less U.S. property CAT because the pricing environment has been severely impacted by lower prices in the reinsurance market.
We've been growing our international property, in particular, the highly engineered property, where we're getting well paid for the expertise that we provide to do the highly engineered property underwriting. We've recruited over 500 engineers over the last three years. That's a very value-added service that makes us a potent competitor to companies like FM Global that are really providing highly engineered property underwriting on an international basis. We've changed our governance in a way that has one person look at our global property business, George Stratts, who can deploy our capacity wherever our clients need it, as opposed to a more historical approach, where the local management would have had their own risk appetite limit our willingness to deploy capital locally based on their local P&L. We now mutualize that globally, which gives us the ability to operate like the large company that we are.
Bottom line, as pricing deteriorates in one market or improves in another, we really are quite conscious about shifting our capital to where it's best rewarded. If it's not being rewarded well at all, the numbers that Josh mentioned earlier in terms of surplus capital, that will go back up, and we'll be looking to buy back more shares. We don't want to just blindly deploy our capital in a softening market, is the simple answer to your question.
I think we're out of time, please join me in giving Peter a hand.
Thank you.
I believe there's lunch available probably about now, too.