Good morning, ladies and gentlemen, and welcome to the second quarter earnings review for Travelers. We ask that you hold all questions until the completion of formal remarks, at which time you will be given instructions for the question and answer session. As a reminder, this conference is being recorded on Thursday, July 22nd, 2010. At this time, I would like to turn the call over to Ms. Gabriella Nawi, Senior Vice President of Investor Relations. Ms. Nawi, you may now begin.
Thank you, Frank. Good morning and welcome to Travelers' discussion of our second quarter 2010 results. Hopefully, all of you have seen our press release, financial supplement, and webcast presentation released earlier this morning. All of these materials can be found on our website at travelers.com under the investor section. Speaking today will be Jay Fishman, Chairman and CEO, Jay Benet, Chief Financial Officer, and Brian MacLean, President and Chief Operating Officer. Other members of senior management are also in the room available for the question and answer period. They will discuss the financial results of our business and the current market environment. They will refer to the webcast presentation as they go through prepared remarks. Then we will open it up for questions. Before I turn it over to Jay, I would like to draw your attention to the explanatory note on page one of the webcast.
Our presentation today includes forward-looking statements. The company cautions investors that any forward-looking statement involves risks and uncertainties and is not a guarantee of future performance. Actual results may differ materially from those projected in the forward-looking statements due to a variety of factors. These factors are described in our earnings press release and in our most recent 10-Q and 10-K filed with the SEC. We do not undertake any obligation to update forward-looking statements. In our remarks or responses to questions, we may mention some non-GAAP financial measures. Reconciliations are included in our recent earnings press release, financial supplement, and other materials that are available in the investor section on our website, travelers.com. With that, here is Jay Fishman.
Thank you, Gabby. Good morning, everyone, and thank you for joining us today. Given the very substantial second quarter weather losses for the entire property casualty industry, and of course for us, we were pleased with our performance this quarter, reporting net income of $1.35 per diluted share, an increase of 6% from last year's quarter, and a return on equity of 10.1%. Recognizing that our operating income of $1.39 per diluted share is about $0.11 below consensus estimates, I'd like to make the following observations. First, the $0.11 differential equates to approximately $50 million after tax. I'd point out that for the first six months of the year, we've earned in excess of $1.3 billion. Secondly, the shortfall is attributable to second quarter weather losses, which aggregated $285 million after tax, or $0.58 per diluted share.
To put that cost in context, the estimate for catastrophes for the second quarter included in our previously provided guidance was $92 million after tax, or $0.19 per diluted share. As best as we can tell, consensus estimates included $0.22 per diluted share for second quarter catastrophe losses. To put the first half weather losses in context, we've already recorded $597 million of after-tax catastrophe losses, or $1.19 per diluted share. The methods used to estimate our expected annual catastrophe losses indicate an expected annual loss of $390 million after tax, or $0.80 per diluted share for the entire 2010 year. Even after just six months, we are now well in excess of our expected annual loss estimate for the full year. Recognizing that we don't control the timing of weather, we just take these events in stride.
There have been periods where catastrophe losses have been exceptionally low, such as in 2006 and 2007, and there are times where they run abnormally high. We price our product for the long term, and weather will occur when it does. We don't believe this high level of catastrophe loss is in any way a result of changed underwriting standards or reinsurance practices, and Brian's going to have more to say about our CAT losses later. Third, the quarter also benefited from $251 million, or $0.51 per diluted share of favorable reserve developments. Again, as best as we can tell, consensus estimates for the quarter included $0.23 per diluted share of favorable reserve development. Just a reminder, because development is so unpredictable, we don't include it in any of our guidance.
In terms of the operating environment for the second quarter, we remain quite pleased with our performance in our Personal Insurance segment, both auto and homeowners. Given the amount of discussion this time last year on our agency auto business, we're particularly pleased with the improving rate of change of policies in force in Personal Insurance. Profitability in the first half improved versus last year, and we're pleased that the actions taken to improve profitability, which we spoke to you about this time last year, are coming through in the results. In our commercial businesses, the operating environment really remained very similar to last quarter. Retention rates remained quite strong and rate on renewal business remained positive, but it was at a lower level than in the first quarter.
The negative impact of the economy on net written premiums has moderated somewhat from recent quarters. We're hopeful that this bodes well for future economic growth. We repurchased $1.4 billion of our common stock in the quarter. Since the second quarter of 2006, we've now repurchased over $12 billion of our common stock. As these actions demonstrate, we continue to execute successfully in the marketplace, generate solid earnings, and return excess capital to our shareholders. Given that municipal bonds have been receiving a lot of attention lately, we thought we'd take a few minutes this morning to have Bill Heyman take you through the strategy and the tactical positioning behind our muni portfolio and demonstrate why we're comfortable with what we own.
Our investment team has applied to our municipal portfolio the same very thoughtful approach to risk and reward that served us exceptionally well during the capital market crises over the last several years. I'm sure you'll find it helpful. With that, let me turn it over to Jay.
Thanks, Jay. There are a few points based on the data contained on pages four through seven of the webcast that I'd like to highlight this quarter. First is our strong balance sheet. All capital leverage and liquidity measures remain at or better than target levels. Second is the repurchase of $1.4 billion of our common shares this quarter and the payment of $173 million in common stock dividends. This brings the total cash we return to our shareholders to over $3.1 billion in the first half of this year. Holding company liquidity of $2.4 billion at the end of the quarter, down as planned from the $3 billion we held at the beginning of the quarter due to our share repurchase activity and the timing of dividends from our operating companies to our holding company. Another quarter in which we increased book value per share.
Operating performance, ex CATs and favorable prior year reserve development, but including net investment income that was in line with our expectations. Another quarter of net favorable reserve development in each of our segments, mostly driven by Business Insurance, where we are now seeing better than expected loss results for property, workers' comp, and commercial auto product lines in recent accident years, and where we have re-estimated unallocated loss adjustment expense reserves given recent loss results. Finally, a double-digit ROE and operating ROE, despite the record second quarter CAT losses. I'd also like to point out that we successfully renewed our CAT treaties this quarter, keeping essentially the same structure, but at a modestly lower cost. The new treaties are outlined on several pages in the webcast and are described more fully in our second quarter 10-Q, which has been filed earlier today.
Let's have Brian now discuss our operating performance.
Thanks, Jay. Before I go into the business-specific results, a few comments on this quarter's catastrophe losses. As Jay mentioned, this was the highest level of second-quarter catastrophe losses in our history. Obviously a difficult weather quarter, and it's on the heels of high first-quarter catastrophe losses. In this quarter, the activity was not of the national headline variety, but it was a very active quarter with numerous significant events. Page eight lists out the 14 industry-designated CATs for the second quarter of 2010 with initial Property Claim Services or PCS estimates. Although 14 second-quarter events is not unusually high for the industry, the severity of a number of these events is clearly out of pattern. The initial industry estimates of $3.6 billion is very preliminary and based on reported activity to date.
This will obviously develop up, and we are confident that once fully developed, this will be one of the largest industry catastrophe second quarters. As we look at our level of losses, we're confident that they are not a result of changes in our mix of business, selection process, geographic footprint or coverage grant, and we don't see anything that would indicate a fundamental change in weather patterns. Slide nine displays our first and second quarter catastrophe loss ratios for the last six years or since our first full year post-merger. It shows that for 2010, both quarters were significantly above average level. In the previous five years, with the exception of the second quarter of 2008, catastrophe weather losses were relatively mild. Accordingly, we don't see 2010 as a trend, but just another data point which we will factor into our assessment of risk and reward.
We're going to stay in the weather business and we'll continue to actively manage our exposure. Now let me shift to discussing the segment results, and I'll start with Business Insurance. When excluding catastrophe unpredictability, underwriting income was very strong, driven by continued favorable prior year reserve development. In the current year, loss trend continued to be fairly benign and consistent with our expectations. The renewal rate change on premium for the first half of the year has been slightly positive, but is below loss trend and less than expected, so the impact on margins is slightly more than we anticipated. Overall, net written premiums were down slightly in the quarter, but not down as much as earlier in the year. The negative impacts of the economy on our insurers, which we see in lower exposures and reduced audit premiums, continued, but have moderated from earlier levels.
Now I'll turn to pricing, and if unpredictability has been the message with the weather, stability is the message in our aggregate domestic commercial insurance pricing. Slide 11 graphs our renewal premium change and splits out the pure rate and exposure change components. The short story here is that nothing is really changing. Overall renewal premium change on guaranteed cost business was essentially flat with a modest improvement in exposure change offsetting the modest decline in rate change. These overall trends are fairly consistent on an individual business basis, but as always, there are differences, and I'd like to highlight a few things. Starting with slide 12 in select accounts. We've been getting some good premium increases, and retentions have been consistent, but a few points below where we historically have been.
New business has been mixed with strong results in smaller accounts. In the fourth quarter 2009 and the first quarter 2010, lower new business levels in Express Plus. In the second quarter, we moderated our pricing strategy and believe this helped boost our Express Plus new business in the quarter. On the renewal book, we typically work 90 days in advance, so any impact on retention will start in the third quarter. Overall, we continue to feel very good about our Select Express platform and how we are positioned in this market. In commercial accounts, retentions have remained very strong, and the renewal premium change has improved modestly, going from a slight negative to a slight positive.
If you look at the commercial accounts chart at the bottom of page 13, you can see that this essentially flat premium change is the result of improving exposure changes and slightly lower rate changes. When negotiating a renewal, we are negotiating both the pure rate and total premium with the account. As the impact of the economy on our customers' exposures has become less negative, it has become increasingly more difficult to negotiate a rate increase, especially when it would result in an overall premium increase. New business results for commercial accounts were down compared to the prior year quarter. In the second quarter of last year, our new business writings benefited from the marketplace disruption caused by the financial distress of several of our competitors. Absent this impact, the results reflect the normal seasonality of our business.
Our new business flow has stabilized at record levels, with quote rates up slightly and close rates down slightly. Given these dynamics, we continue to be very pleased with our new business performance. In other Business Insurance, retentions have remained very strong, while the price change for both the rate and exposure have moved more significantly than the overall commercial business. These pricing impacts are primarily driven by our large property business, with some significant softening in rate but improvement in exposure. As competitive pressures in this class of business increased in the current quarter, driven by ample reinsurance capacity and a benign 2009 hurricane season, we made a conscious decision to maximize our retention on this business and not push rate on large property accounts.
Similarly, large property new business became increasingly more competitive, with aggressive concessions to expiring terms and conditions frequently required to motivate accounts to leave their current carrier. We are not willing to make these concessions. In summary, we continue to prudently manage each of our businesses and adjust our actions to the unique needs. In the aggregate, the results are not different than our recent experience. That is solid retention, relatively stable pricing, and a strong flow of new business opportunities. In the Financial, Professional & International Insurance segment, the core underwriting margins were generally stable in the quarter as the marginal rate gains essentially offset slightly increasing loss trends. Additionally, the underwriting margins benefited from a reduction in surety reinsurance costs associated with prior year reinsurance treaties. Net written premiums, after adjusting for the impact of changes in foreign exchange rates, were down for the quarter.
In some of our management liability lines of business and in some of our international lines, we believe pricing is not consistent with our profitability targets. Our writings in those lines are down. Turning to the production statistics on page 16, although renewal premium change is slightly negative in both management liability and international, for most of the businesses contained here, we've achieved positive rate gains. Turning to Personal Insurance, in agency auto, our underwriting margins, after adjusting for the impact of CATs and prior-year development, improved quarter-over-quarter as rate gains once again outpaced loss cost trends. Within agency property, we have been speaking to you for some time regarding the slight margin compression driven by year-over-year increases in the cost of materials, primarily asphalt shingles.
As the rate gains we achieved in recent quarters continue to earn through our property business, we have crossed an inflection point where in the second quarter our rate improvement outpaced loss cost trends. Agency auto new business improved in the second quarter as we reintroduced our 12-month policy. As we've become more confident in our new business products and are consistently within target returns, we felt it was appropriate to reintroduce a longer-term policy. The aggregate timing impact of this additional six months of written premium was about $30 million in the quarter. As the rollout of this product is completed, the impact on our new business written premiums will increase throughout 2010. More significantly, we are pleased with the improving rate of change in policies in force. We believe the PIP change is a result of our improved competitive position compared to third and fourth quarter of 2009.
Agency property production results for the quarter continued to be strong in spite of the difficult housing market. Quarter-over-quarter PIP growth was at the highest level since third quarter 2007. Retention improved two points, and new business is up 15% compared to the same quarter last year. Given the expansion in core underwriting margins and the strong top-line trends in these products, we are very pleased with both our current and going-forward marketplace position in Personal Insurance. Let me sum it up. Weather's been bad, but it's a big part of our business, and the one thing we know for sure is that it will change. The commercial marketplace, both domestic and international, is fairly stable. We obviously wish it was improving, but we feel great about our position and are confident we will maximize the opportunity.
In Personal Insurance, we are encouraged by the trends both for us and the industry. With that, let me now turn it over to Bill Heyman to discuss our tax-exempt portfolio.
Thanks, Brian. Since municipal securities have received considerable attention in the press, I thought I would spend a few minutes discussing the sector in general and our portfolio in particular. As you know, we have a tax-exempt portfolio of about $41 billion. You may not know that of that, about $7 billion consists of bonds which have been pre-refunded, which means they have been defeased to maturity or their first call date, usually with U.S. Treasury securities. Still, the $34 billion remaining constitutes about half of our fixed income portfolio. The first observation I'd make is that we have, fortunately, always, starting long before credit quality became an issue, viewed tax-exempt credits as credits first, without regard to any advantages their tax-exempt status confers.
While optimization of the alternative minimum tax provides a target allocation, it's not a bucket we feel we must fill regardless of qualities. A general obligation of a school district competes with a corporate bond or a mortgage-backed security for our investment dollars. We scrutinize closely not only the creditworthiness of the issuer, but the nature of the obligation. For example, in some cases, revenue bonds could be stronger than GOs and vice versa. Furthermore, where issuers have enhanced their creditworthiness with bond insurance, we have always ignored it. The result was that several years ago, when the bond insurers more or less simultaneously lost their AAA ratings, we did not make portfolio changes. The result of high credit underwriting standards means that our portfolio does not closely resemble the broad market.
For example, we own about $23 million of municipal healthcare revenue bonds, or about one-tenth of 1% of our portfolio. Such bonds represent roughly 7% of the broad market, and since 1970, 39% of all municipal defaults have been in the healthcare sector. Conversely, we are overweighted in pre-refunded bonds, state and local general obligation bonds, and water and sewer bonds. Recent focus on general obligation bonds relates to the potential for these bonds to be treated like senior unsecured debt, with bondholders becoming general creditors alongside employees, retirees, and vendors. The general creditor construct is most descriptive of general obligation bonds issued by states, as their general funds act like pools, collecting revenues from a wide range of taxes, fees, and paying salaries, making pension contributions, paying vendors, as well as servicing their general obligation debt.
In contrast, this construct is probably not descriptive of many local general obligation bonds. We expect that in extremis, holders of many local GOs would be treated as secured creditors based upon legal precedent and the provisions of Chapter 9 of the Bankruptcy Code, which provides that the security interest in special revenues remains valid and enforceable in bankruptcy. As you know, most school district general obligation bonds are specifically authorized by voters in a ballot initiative. This authorization allows the pledge of ad valorem taxes to be levied on all taxable property within the school district without limitation as to rate or amount. The collection of these taxes is generally segregated in a debt service fund outside of the general fund. These funds can only be used to service the authorized debt and are not available to finance the general purposes of the school.
The unlimited nature and segregation of this debt service tax revenue contrasts sharply with the school district's other property tax revenues, which are statutorily limited and commingled with its state per-pupil funding and used for operations. In addition to assembling the portfolio selectively, we manage it actively. The portfolio contains approximately 8,000 discrete securities, but of only 925 issuers. There are over 50,000 issuers in the municipal market. We have general obligation bonds of a mere 75 cities. There are 5,300 in the Bloomberg database. We feel we know our cities well. We monitor closely the financial condition and operations of our credits. While it is true that we are basically buy and hold investors, we have in the past year reduced, often substantially, positions in issuers whose creditworthiness has, in our view, deteriorated.
I would add, we believe that even those issues we sold are overwhelmingly likely to pay every dollar on time. As we demonstrated two years ago with our mortgage portfolio, we do not blindly rely on ratings. It is nonetheless worth noting that roughly 50% of our municipal portfolio holdings, which are not pre-refunded, carry an average AAA rating from Moody's, S&P, and Fitch, compared to only 17% of the broad market. Almost 96% are rated Aa3 or higher. Only 4%, 4.3%, are rated A or lower, compared to 33% in the index. Rating agencies are looking at the same challenging economic, budget, and spending dynamics affecting state and local governments that all of us are. Yet the ratings of our holdings remain very strong.
A recent publication by Moody's points out that over the entire period, 1970 to 2009, the riskiest 20% of municipal issuers, measured by ratings, accounted for 86% of all defaulters. Our portfolio contains approximately $7.2 billion in state general obligations, with only about $330 million in the headline states of California, Illinois, and New York. The table on the webcast contains the aggregate holdings and the average ratings of our top 16 states. With the exception of our holdings from California, which average Aa2, the average rating for each state is Aa1 or higher. The general obligation bonds of the top five states total $1.4 billion. All that said, we have purchased very few big state GOs in recent years and have sold a few recently. The other slide of the webcast presents the runoff of our state GOs.
Based on current market yields for those borrowers and the coupon structure of our holdings, nearly 70% of our state GOs will mature or be called by 2015. The local general obligation bonds from the top five states total $6.4 billion, and the revenue bonds from these states total $3.6 billion. Our general obligation bonds of issuers other than states consist largely of school district exposures with a moderate number of counties, cities, community college districts, and other special districts. As we just discussed, the nature of these obligations is such that most could be considered as obligations secured by a pledge of revenues from ad valorem property taxes levied without limitation as to rate or amount. As we have discussed previously, our fixed income portfolio exists first and foremost to ensure our commitment to our policyholders.
We underwrite the invested risks assumed in our portfolios with careful consideration of the trade-off between risk and return. Just as in the other fixed income assets, we do not reach for yield in our municipal portfolio in an attempt to outperform a benchmark, boost investment managers' compensation, or for any other reason. We expect our fixed income assets, including our municipal portfolio, to provide adequate risk-adjusted returns and support our insurance operations in pursuit of maximizing wealth for shareholders over the long term. Obviously, history may not be indicative of the future, and ratings cannot be relied upon with certainty. We believe our portfolio is strong, and we wouldn't trade it for anyone else's. We could certainly have losses in our municipal portfolio, maybe even material losses. Even viewed through the prism of recent events, we feel very comfortable with what we own.
Let me turn it over now to Jay Benet.
Thanks, Bill. Let's discuss updated guidance for the full year. We've got information on page 22 for full-year 2010 fully diluted operating income per share, which has now been refined from the previous range of $5.20-$5.55 to a new range of $5.20-$5.45, which is a $0.10 reduction of the upper end of the range that primarily resulted from commercial renewal premium increases in 2010 not meeting our original expectations, which we believe is attributable to the impact of the continuing difficult economic environment. In round numbers, this range for operating income should still translate into an operating return on equity of approximately 11%, as we've said before.
We continue to anticipate some accident year loss ratio deterioration on a consolidated basis for full-year 2010 ex CATs, as we expect loss cost increases to modestly outpace projected earned rate increases in our commercial businesses for the full year. In your modeling, please remember that the second half of last year included favorable re-estimations of current year loss ratios for first half 2009 losses. Quarterly loss ratio comparisons for 2010 versus 2009 must take this into consideration. We're now assuming CAT losses of $835 million after tax or $1.71 per diluted share, which incorporates our actual CAT losses for the first half and our original estimates for the second half of the year. No further estimates of prior year reserve development, either favorable or unfavorable.
A low single-digit decrease in average invested assets, ex unrealized gains and losses, resulting from a reduction of holding company liquidity due to the share repurchases. A full-year share repurchases of $4 billion and a weighted average diluted share count after share repurchases and employee equity awards of approximately 487 million shares. With that, why don't we open it up for Q&A?
Thank you. Ladies and gentlemen, if you would like to register a question, please press the one followed by the four on your telephone. You will hear a three-tone prompt to acknowledge your request. If your question has been answered and you would like to withdraw your registration, please press the one followed by the three. If you're using a speakerphone, please lift your handset before entering your request. One moment, please, for the first question. Our first question comes from the line of Jay Gelb from Barclays Capital. Please proceed.
Thanks. My first question is on the pace of the share buybacks. Travelers repurchased $3 billion of stock in the first half of this year. I believe the guidance is for $4 billion for the first year, yet there's still almost $4 billion left in the existing authorization. I'm just trying to get a sense of why the pace may slow, or might that just be a conservative outlook? My second question has to do with the economy. It appears that we're lapping some of the more challenging comparisons. I'm trying to get a sense of how much of a benefit the stabilizing economy could be and whether Travelers feels it can still raise prices as exposures increase. Thanks.
Jay, thank you. First on the share buyback. We have historically scaled back in the third quarter, pending, obviously, cat season. Of course, it seems to us like we've already endured two quarters of cat season already. My recollection was that the original estimates of share buyback included in the guidance was $3 billion-$4 billion.
3.5%-4%, 3.5%-4%. Essentially, we are still largely on that target for the moment. We may alter that as we go into the quarter, depending upon the second half of the year, that is, depending upon earnings, depending upon weather, reserve development to the extent it occurs, and all that. Nonetheless, it's a plan that we embraced at the beginning of the year, and for the moment, we're sticking to it. Your second question, and maybe Brian and I will ham and egg this a little bit. We certainly can't prove this to you. It's largely anecdotal. It's from sitting in with conversations with our field folks, and I'm speaking now about your question regarding the commercial insurance environment. Our sense is that there isn't anything that has particularly changed competitively. You all seem to be fixated all the time.
Have some company or companies changed their pricing strategy or tactics? We certainly don't see that and have no evidence of it. What we are hearing from our field folks is that as exposure is leveling out, and a comment I made earlier was that it looks as though exposure is trying to get back to zero, broadly speaking, in the commercial lines businesses, is that it's getting more challenging to get rate increases. If you're a customer, you really don't focus on the granular dynamics that make up your price, your premium. You're really not all that worried about how much is exposure and how much is rate change and all that nonsense. What you, as a customer, tend to look at is my premium going up or down in its simplest form.
As exposure flattens out to the extent that you try and get rate gains, you're asking for a premium increase. In this economic environment, our sense is that it's getting just more challenging to pass that rate increase on. There's a little more resistance both from the intermediaries, the agency brokers, as well as the customers to accepting meaningful rate increases. Now, it's interesting. We didn't present the data, but we certainly do look at the distribution of rate gains in our middle market business each and every quarter. We've shared that with you previously. To the extent there are loss dynamics at work, to the extent that an account has loss experience that's different, more problematic than what it had originally been expected, we do get rate gains.
There are still significant accounts where you'll see a +5 or even a +10 rate gain, but it largely is loss driven. It's not, in effect, margin or profitability driven. I'd love to if Brian has any comment, but our sense is that the margin picture broadly in our commercial businesses is modestly deteriorating. That is that our ability to get rate in the aggregate is a little bit less than what you're experiencing, broadly speaking, in loss trends. That's the way that I think I'd answer it. I don't know if you have anything-
I mean, one of the phrases that I'd emphasize that Jay just used is this economic environment. We certainly wouldn't want to make the statement that as long as exposures are coming up, you can never get rate changes. If you think of where the psyche of our typical commercial customer is today, they saw the economy take a big dip. They saw their business, by and large, take a big dip. The good news is, if we look at our total portfolio, they're kind of back to not dipping anymore. They're not yet back to growing or anything close to robustly growing. In that mindset, they're sitting there going, again, loss experience being neutral, we're going to struggle with premium increases. That's the dynamic we're seeing today.
In the context of returns in our business, you always got to go back to where are we from a profit margin perspective. If you go back to the guidance that we gave at the beginning of the year, which obviously contemplated normal tax and no reserve development, the return on equity that we were projecting for the year was about 11%. In the context of the environment we're in, both the investment environment, important to remember is the investment environment combined with the economic environment, that's a pretty good return. It's difficult for us to plead that profitability measures in our business are inadequate, and as a consequence, we need to improve margins. In this investment and economic environment, to the extent that we're capable of producing an 11% return on equity, we feel pretty good.
We continue to try and seek rate where most importantly, where the loss experience demonstrates that we should. To the extent that we can improve our profitability dynamic as we move forward, we'll continue to do that. It's not as if we're starting from a point of impaired profitability in any way. It's a pretty strong franchise and a pretty strong profit picture in the context of that environment.
That's great. Thank you.
Pleasure.
Our next question comes from the line of Matt Heimermann from J.P. Morgan. Please proceed.
Hi. Good morning, everybody. Couple questions for Bill. One, I was just curious if you could comment on the performance of the other investment income line this quarter and what drove that. Just a little surprising given what macro was happening in the investment environment. Second, just with respect to the municipal portfolio, I guess if someone wanted to play devil's advocate, a lot of the defenses of the municipal securities market around rating, average credit quality, things like that, were things that were made for the mortgage market. I guess, if you had to be realistically bearish on your portfolio or the market broadly, what would you point to as kind of the bigger risk?
Okay, let's take them in order. In terms of the other assets, the results of the quarter were attributable to strong distributions in private equity. Cash flow from the portfolio for the first six months was about even. That means distributions equaled capital contributions, which is better than we would have expected, and there was a lot of net investment income gain in the results that we saw. Hedge funds made money, but not very much. Maybe a couple points better than the broad index. Our hedge fund portfolio was only about $470 million. It was $1.6 billion at the time of our merger. Frankly, we find it difficult to predict ex ante facto which managers will do well and which won't. We have a handful of relationships of long-standing.
In real estate, a lot of our funds had marked down properties probably excessively, and the first half of the year reflected a balance in valuations. Let's go to municipals. I agree with you. One can argue both sides of the equation forcefully. We think about this issue every day. I would add, we're not talking our book because we can't get out of where we are. If we decided that instead of $34 billion, forget the pre-refundeds for the moment. Instead of $34 billion, it ought to be $20 billion or $25 billion. The quality of our holdings would permit us to make that adjustment. The market is very firm. Frankly, it's very firm, even in credits we wouldn't buy. In terms of the risks in the sector, I think I alluded to this at Investor Day in May.
I think the risks are less of economic default because even some states with big budget problems are, when you analyze them, not very heavily taxed. The risks are that some jurisdictions will seek, if a court allows them, it may not, to repudiate their debt simply because they choose not to increase taxes or cut budgets. I think with respect to larger jurisdictions, the possibility of this is pretty remote. It is one thing to only have access to the capital markets at a high price, and frankly, most jurisdictions haven't even faced that penalty yet. It is another to lose access to the capital markets altogether which is what would happen in a political repudiation. I agree, the issue is not free from doubt, and our take on it can be disputed. It's not a ride that light is.
Okay. Much appreciated. Thanks.
Our next question comes from the line of Keith Walsh from Citigroup. Please proceed.
Good morning, everybody. Couple questions here just to follow up on the exposures. Trend looks better. How do we reconcile that with sort of the negative data that we've been seeing coming out of small businesses, whether it's lending, business optimism, new business startups? If you could also touch on the trend you're seeing within audit premiums, and then I've got a follow-up for Bill. Thanks.
I'll take the first, I'll ask Greg to take on the second. First, I think our exposure data is actually largely in sync with what we see as to GDP changes broadly. We're not seeing growth. What we're actually seeing is a leveling out, and I think that there are some measures that would suggest that the economy is actually expanding modestly. There are some that would suggest that it is flattened, and that's largely, at this point, what our aggregate data says. There'll be some pluses and some minuses in various businesses. The exposure change that we saw in the second quarter across all of our commercial businesses in the aggregate is getting close to zero. I think if you asked anyone from an economic perspective, what do they see? They would largely say that at the very least, we've flattened out.
We don't see it as inconsistent. Maybe there's something else that you see, that's kind of our take on it.
Keith, on the second piece, did you ask about auto or audit?
Audit. Audit premiums.
Audit. Okay.
Oh.
Okay. This is Brian. On the audit premium side, we've talked about the numbers before. In a normal environment, we're going to see something like three to five points of positive on the audit premium side. Around the end of last year, very beginning of this year, we were running about one point negative, and we're close to back to zero. We're at a really modest negative number. Pretty consistent with that same view on the exposure change numbers. Still negative, but close to back to a plus number. Now, again, in a normal world, we'd look for a plus three to five.
That's less about, do you want to speak to that point about it being less about the economy than the way our policies work?
Well, yeah. There's obviously the audit premium has a lot to do with the psychology of the buyer and how the product is even sold. The natural course is that the buyer is going to understate exposure slightly, and we pick it up on the back end on the exposure. We're always playing that game. That leads to the normal three to five.
that one may snap back.
Quicker
Somewhat unpredictable because we're dealing with a psychology of our buyers and their willingness, in effect, to have us carry for a period of time, premium relative to exposure change. We, of course, factor that into our pricing. It's not a surprise to us. That one may snap back faster because it is predominantly psychologically driven more than the overall genuine exposure, which really is an economic dynamic.
That's very helpful. Bill, you started to touch on this on the munis, what I really want to know is who holds the senior position here? Is it the bondholders or is it muni workers, pension payments, and healthcare benefits? Where's that dollar going to go to if it has to, if we come down to that?
Well, we think in issuers below the state level, bondholders. At the state level, less clear.
Okay, thanks.
Our next question comes from the line of Michael Nannizzi from Oppenheimer. Please proceed.
Thank you. Just had a question about the FP&II segment and reserve development there. Can you talk about that development relative to your management liability business, in particular, the depository institutions? I have one follow-up. Thanks.
Sure. I guess I'd direct you back to the conversations we've had previously about the credit crunch analysis, and I'd say that broadly, the way that we're looking at it is stable. It hasn't been a lot of change. The release broadly comes from six or eight years, from 2001 to 2008. The actuaries go through the data every quarter, and they are making lots of estimates and judgments based on a lot of cases that are still really in a state of infancy. I would say broadly speaking, no change.
Okay. I know that disclosure last year, you had mentioned 26 institutions that had been closed. Can you update on the exposure there and any changes or any new institutions?
You mean specifically on the institutions that have been taken over?
Yes.
Yeah. So far this year, I think we've got probably close to 100 institutions that have been taken over, about 96. We've got 33 of those.
At first you said, "We've got." There have been. Sorry.
I'm sorry. There have been 96 institutions the FDIC has taken over.
Right.
We're on about 33 of those. Our average exposed limits to those are about $4.5 million. Broadly speaking, it's developed since last year as we would've expected it to.
Great. Thank you. Just if I could one follow-up on personal lines. Brian, you'd mentioned that slide had $3.6 billion, I think, in CAT expectations for the second quarter, and Travelers' number in the second quarter is about $440 million. Am I looking at that right? I mean, that's about 10%. I just want to understand, is it just different estimate base or how should we think about that? Thanks.
The $3.6 billion on the industry slide comes out of PCS, Property Claim Services, and that is purely the kind of initial data that they've gotten from carriers. And in some cases, for some events that happened literally three or four weeks ago.
Right
very, very immature data. That will develop up fairly dramatically.
History would suggest.
History would suggest, yeah. Just the way the claim process works.
Could I make a very important point? We don't utilize PCS estimates to make our estimates of losses from those events. Our losses are based upon what we're accruing, are based upon claim notices, claims filed, and our people on the ground. The important point here is that while we believe that history, again, suggests the PCS data will develop negatively, historically, our estimates have actually been pretty good relative to catastrophe claims.
Yeah, because obviously what we're trying to do is, as quickly as possible, take our claim data and make an actuarial estimate of what the ultimate will be. PCS does that over time, but it takes them longer. It's really tough to do any kind of industry relativity right now. Then obviously you've got geographic distribution underneath that you got to look at. We will scrutinize that as the data becomes available. All we know is it's a big industry event and it was a big quarter for us, too.
Great. Thank you for answering my questions.
Our next question comes from Jay Cohen, Bank of America, Merrill Lynch. Please proceed.
Yes, thank you. A couple of questions. I guess first, maybe big picture on the reserve development. I guess if you looked at your accuracy, you'd almost think you guys aren't that good at setting reserves given how redundant things have been. If you can talk more specifically about what's happening. I know it's going to vary by line and by segment, but is it more of a frequency issue or a severity issue? That's the first question. Then secondly, maybe for Bill, on the other investments, it looks like the return this quarter annualized was around 12%. I probably asked this before, but what would you consider the normalized return on those investments?
Jay, Brian is going to answer your question on the reserve development, I would just make an observation that we really try hard to get it right, because the issue is actually how we price our product. To the extent that we end up overstating our loss costs going into a product offering, then we are pricing it higher than we otherwise would have to, and we become less competitive relative to someone who takes a different view. It always feels good when it happens, of course. It feels terrific when your results are impacted this way, and certainly, it is a whole lot better than underestimating your loss cost, which means you are underpricing your product, and that is really a disaster.
We really try hard to get it right, because in the end, what we are about is trying to grow our business, and getting it right is what is going to help us do that the best. Now, your point is relevant. We have had a long stretch here of significant development. Brian or Jay?
Yeah, I can start. I think when you look at the components that make up the reserve development, frequency continues to be behaving in a way that when you go through the reserve setting process, you come up with estimates of what frequency is going to be, and it is behaving better than that. That has been a trend that has been going on for a while. When actuaries and finance people are looking at what the reserve process should be as it relates to favorable frequency and work with their business partners in terms of pricing and everything, if you look at what the current trends are, you wonder, are things going to get better than they are when you are in a favorable position? Sometimes you just look at it and say, "Well, they are pretty good, and I cannot imagine them getting better," and sometimes they do.
That has been a part of it. We are in a very low inflationary environment today. Looking out several years as to what severity is going to be, just a little bit of a change in the inflationary outlook is going to have an impact as well. You have to go back to what is the base upon which all of this is being done. On a gross basis, our reserves are in excess of $50 billion. The absolute level of these changes in terms of percentage points are very low, but given the nature of that reserve base, they impact the income statement as you have seen. As Jay said, each one of these quarters, we are doing best estimates of what we see out there using all the information, and as it changes, we just update the reserves.
Jay, this is Brian. Two other points. If you look at where the development has come from, not just in this quarter, but over a good period of time, I'd say two things. First, on the liability side, where we've had a lot of favorable reserve development, one of the toughest things to guesstimate at over time is where the tort environment's going and what the sustainable impact of that is. That's been a big variable driving our improvement. At the same time, over, and again, this is a five to six year kind of statement, we've done a lot in our liability claim process that we think has helped the way we do business and service our customers, and we think that's had a pretty strong impact there.
It takes time to see the impact of those things coming through the loss data, that's been driving it. The other area is workers' comp, which has continued, again, I always say this with comp, very much a state-by-state game. In the aggregate where we play, we've continued to see some favorable moves there and frequency dynamics that Jay Benet talked about. Comp would be an area where that's really been true, is trying to look at those frequency trends and project forward what do we think they're going to be in the future.
That's great. The other investment return?
Yeah. I think that one has to speak to various sectors rather than generalize for other investments. I think from owned real estate, unleveraged, one ought to hope to make mid to high single digits. Leveraged real estate, that is funds, low double digits. Hedge funds, low double digits. Private equity, maybe a little more. Our blended return depends on our allocation of those asset classes. Obviously, one has to rethink all those returns if we are in a protracted period of very low interest rates. We may all of us, still be stuck in the mindset of what these asset classes ought to make in the world before 2008. Those are horseback numbers.
Based on what you just said, the returns in this quarter weren't terribly overstated at all.
They are generalizing as a blended number. They're what we would hope to make over a statistically significant time period, which unfortunately is more than a quarter.
Got it. Thanks, Bill.
Yeah.
Our next question comes from the line of Cliff Gallant from KBW. Please proceed.
Good morning. Building actually up on Jay's question a little bit, in terms of loss cost trends, I'm just wondering about to what degree are you seeing favorable loss cost trends, whether it be from deflation or even falling frequency? To the extent that you are, how confident would you be to really act upon that in terms of your forward pricing? I think of the Personal Auto industry and 10 years ago, where they were about to experience several years of rapidly falling frequency. I think some of the more aggressive companies like Progressive had wished that they had been able to fully anticipate that to really grow their book. I'm curious how you would take that in terms of going forward.
Well, first, I'd observe that this is very much line-by-line specific. There are long-tail lines where you take a deep breath, obviously before you embrace an emerging trend. I'd make an example in workers' comp. We are still contemplating meaningful medical inflation. We're not pricing our product. We talk all the time about benign loss costs and no inflation. We should always be excluding medical work and medical portion of workers' comp, where we assume a robust inflation rate, and we'll obviously see how that works out. Generally speaking, with the, I think a couple of individual exceptions, Brian talked about asphalt shingles, which a number of carriers have highlighted because it's a byproduct of the price of oil, that loss cost trends severity have, broadly speaking, been awfully benign. That doesn't mean zero. It doesn't mean that we don't see some embedded inflation in lines of business.
The shorter the tail of that business is, the more we're willing to embrace it and incorporate it into our pricing, because if we're wrong, it doesn't take long before we can reprice the product. The same thing holds true for frequency. In many lines of business, there has been a long-term systemic decline in frequency. I would actually say that well, we've been checking it. You ask the question all the time, as frequency comes down, is it making a new level? Is it going to return back to previous levels, not only from a pricing perspective, but also from a reserving perspective as well? I think the same thing holds true to the extent that these are shorter tail lines of business. We're pretty aggressive in embracing it.
To the extent that they're longer tail, we're a little more reluctant to move in that direction. I think one of the things that's affected workers' compensation, I can't prove this, it's just from data that we look at, is that as our economy has moved increasingly to a service-based environment, frequency in workers' comp has come down. That's not a great surprise to people. That's sort of the dynamic, that element that you feel more comfortable embracing. It's very line specific, and I think it's difficult to generalize. If there's a line that you're particularly interested in, we can attempt to answer.
Cliff, you're really getting to the core of the puzzle we're trying to solve each and every day in the marketplace. Do we understand why loss trends are changing? First of all, can we see as granular as possible where they're changing and breaking out? The real trick is understanding why. The generality would be the better we feel about our ability to understand why something's changing, the more confident we're going to be factoring it into our pricing. Where you get in trouble is you see a pattern, and you don't understand why, and you just start extending the pattern. We could talk. It is, as Jay said, it's very much a line-by-line.
I do think, though, culturally, as a matter of the institution and the way actuaries think and in this organization, you would characterize us as not pushing the edge. Again, particularly in longer tail lines, you take a great risk when you make an assumption as to changing trends, embrace it, price your product, reserve for it, and find lo and behold, that you've not only underpriced but under-reserved both. I just think culturally here, we're not an organization that instinctively pushes the edge.
Okay. No, it's a fair answer. Thank you.
Our next question comes from the line of Brian Meredith from UBS.
Please go ahead.
Good morning, everybody. A couple quick questions here. First, looking at the Business Insurance accident year combined ratios, ex-CAT underlying, particularly on the loss ratio side, it looks like when you adjust for the current period development last year, it's up about 200 basis points year-over-year. Is that pretty consistent with what we should expect here going forward, given your comments about continued accident year deterioration year-over-year?
Brian, I'm trying to do the math in my head.
I think you said it was about 62 last year in the second quarter ex the current period.
Yeah. I would say pretty close to that.
This is Jay Benet. The thing that we'd urge everyone to do is look at last year's combined ratio, look at last year's loss ratio component of that in its totality for the year because of the re-estimation that took place in the third and fourth quarter. The best proxy for what the loss ratio was for coming into 2010 would be the full year 2009 loss ratio. Then we said we would expect some margin compression from that. So far what you've seen is the first six months. That would be the mechanical comparison that you would want to do, recognizing that in any period, the non-CAT related weather is also going to bounce around and there's going to be some noise from other large loss activity or whatever. That would be the base information to be taking a look at.
Great. Then also in the Business Insurance, your G&A expenses, even if you add back the benefit you had in the second quarter last year, down fairly substantially year-over-year. Anything unusual going on there, or is there some expense initiatives going on?
Well, we're always closely managing expenses. There is the timing of some expenses that'll take place at various quarters, one year versus the next. I wouldn't say there's anything in particular going on there other than we're always watchful of our expenses.
Right. The last question, Jay, you talked about pricing and how you want to adjust it for obviously what's going on with loss cost, et cetera, what about interest rates right now? It seems like interest rates are going to be low here for a while longer. How can you adjust pricing for interest rates, or can you in the current marketplace?
If you had gone back a year ago, our view was is that the investment environment that we were in was shorter rather than longer, that it was likely that the combination of stimulus dollars and whatever else, the Fed ultimately backing away from the support of the financial institutions, that rates would rise to more normal historical levels. Bill and I spent a lot of time talking about the environment, it's very possible now that what we're in is an extended period of time at where we are. As a consequence, we are beginning to think meaningfully about what the implication of that is for our business. That not only gets to pricing of our product, but return expectations, capital embedded in our business, ways in which the capital is used. It's a broad question that gets well beyond, are you pricing adequately?
Again, in this environment, with these interest rates, we're able to generate round numbers, an 11% return on equity. You begin to ask yourself the question, if riskless rates are where they are and spreads are where they are, what are reasonable returns and what can be expected, how do we deploy our capital forward appropriately in that very different environment? We're at the, I would say, the early stages of those discussions, they are serious discussions that we're taking place here regularly.
Great. Thank you.
Our next question comes from the line of Paul Newsome from Sandler O'Neill. Please proceed.
Good morning, thank you for the call. First question, I think I accurately heard that you are issuing annual policies for your Personal Insurance. If that's the case, I'm a little bit curious as to the strategy. Historically, I think of that as sort of a top-of-the-market kind of strategy to hold onto your business as opposed to a bottom-of-the-market strategy given that on the upside, you would typically want to be able to reprice your product as fast as you can. Am I just wrong on that, or is there a different subtlety here to that strategy that I'm missing?
Hey, Paul, this is Greg Toczydlowski. I guess to answer that question, you got to kind of start off from where we began. We started driving a sophisticated product out in the marketplace about five years ago, and underneath that was our first foray into predictive modeling. As we needed to stay nimble and adjust that product, we kept the shorter-term contracts. As we move forward and become very confident in that new business pricing, that in combination with the demand base from our agents and customers based on the profile that we're trying to attract, we think it was prudent to issue the annual policy at this point in time. Again, as Jay talked about some of the underlying dynamics of the Personal Insurance business, we feel terrific about that position right now.
My second question is back to the current returns. I agree that 11% ROE, especially relative to your peers, is a good result. The question I have is that does this, in your mind, suggest that essentially the market, meaning the property casualty market, largely ignores the interest rates than where they are today in their profitability? Is this, in your view, a view that essentially the hurdle rate has fallen for the industry? I would imagine that the hurdle rate, if anything, has gone up given volatility, uncertainty, the CAT losses, the stock market seems to be behaving that way. The stock market looks like it's behaving as if the hurdle rate for the industry has risen over the cycle. Any thoughts?
I understand the question, I think, anyway. First, I certainly don't think that the insurance industry ignores investment returns. As we price our product, we are driven by available investment returns in the marketplace. Shortly after the crisis, when the world sort of settled back down and we were confronted with a different kind of fixed income environment than had existed previously, the real question was, were you going to knee-jerk react to a changing interest rate environment, causing disruption to agents, brokers, customers, and all of your distribution, or were you going to take a deep breath and see what would happen to investment returns over time? We've spoken about this earlier. It was our decision to take this in the longer view and not to knee-jerk now, 18 months ago, and attempt to drive pricing up in some very dramatic fashion.
I think that would have been not only highly disruptive, but candidly not very successful. That would have been a strategy that I think would have failed. Now, I think, and again, this is how we perceive it, you can certainly ask other companies what their view is. It may very well be that this economic malaise, such as it is longer rather than shorter, and it may very well be that this investment environment is, as a consequence, longer rather than shorter. We are in the, as I said, the early stages of really thinking through the implications of that. It's a long view. We speak about returns over time. We try hard not to get tangled up in, candidly, this quarter, next quarter, or that we manage this business to generate returns over time.
One of the questions that we're going to ask ourselves is, in the investment environment that's available and given other investments, we recognize that we compete against other investments, including investments that are fixed income. We have round numbers of 3% yield at the moment. What are the implications of that? How do we think about our company, and what do we strive to produce for investors in a very different investment environment than existed even a couple of years ago? It's a very legitimate question, and we're ankle-deep in it at the moment as we try and wade our way through it. It's complex. Obviously, no one company can adopt a strategy that's disruptive in the marketplace and think somehow that it will be successful. We operate in a very competitive environment, so you have to balance all of those factors. More to follow.
Just no immediate clear answers about that issue, but a very relevant question.
Great. Thank you very much.
Ms. Nawi, I will now turn the call back to you. Please continue with your presentation or closing remarks.
Okay. Well, thank you very much for joining us this morning. If anybody has any additional questions, you can reach either myself or Andrew Carlson in the investor relations department. Thank you, and have a good day.
Ladies and gentlemen