American Express Company (AXP) Q3 2011 Earnings Report, Transcript and Summary

Ladies and gentlemen, thank you for standing by. Welcome to the American Express Third Quarter 2011 Earnings Release Conference Call. [Operator Instructions] As a reminder, this call is being recorded. I would now like to turn the conference over to our host, Mr. Rick Petrino. Please go ahead.

Thank you. Good evening. We appreciate everyone joining us for today's discussion. Before I turn it over to our CFO, Dan Henry, I do need to remind you that the discussion today contains certain forward-looking statements about the company's future financial performance and business prospects, which are subject to risks and uncertainties and speak only as of today. The words believe, expect, anticipate, estimate, optimistic, intend, plan, aim, will, should, could, likely and similar expressions are intended to identify forward-looking statements. Factors that could cause actual results to differ materially from these forward-looking statements, including the company's financial and other goals, are set forth within today's earnings press release, which was filed in an 8-K report and in the company's 2010 10-K report, already on file with the SEC. In the third quarter 2011 earnings release and earnings supplement, as well as the presentation slides, all of which are now listed on our website at ir.americanexpress.com, we have provided information that describes certain non-GAAP financial measures used by the company and the comparable GAAP financial information. We encourage you to review that information in conjunction with today's discussion. Today's discussion will begin with Dan Henry, Executive Vice President and Chief Financial Officer, who will review some of -- some key points related to the quarter's earnings through the series of slides included with the earnings documents distributed, and provide some brief summary comments. Once Dan completes his remarks, we will open the line for Q&A. With that, let me turn the discussion over to Dan.

Okay, thanks, Rick. I'll start on Slide 2, Summary of Financial Performance. The revenues net of interest expense was $7.6 billion, 9% higher than the prior year. Now this growth rate is down slightly from the second quarter, but it is best-in-class compared to issuing competitors. On an FX-adjusted basis, it's 6% growth. Net income came in at $1.2 billion, 13% higher than last year, and EPS is $1.03. Return on equity is 28% and you can see that shares outstanding have decreased and that's a result of share repurchases, which I'll cover in more detail later. We move to Slide 3, Metric Performance. You can see the Billed business was $207.7 billion, up 16% and 13% on an FX-adjusted basis. So Billed business, again, showed strong growth. It's the seventh quarter in a row that we've had strong Billed business growth. We are growing over strong growth last year, while our competitors' spending growth last year -- while improving or growing the overall lower growth rates the year before. Cards-in-force are 95.8 million cards, up 8%. That represents 21% growth in cards in GNS and 2% growth in proprietary cards. Average spend continues to increase, reflecting the high level of engagement of our Cardmember base. Cardmember Loans are up 2%, the same growth rate that we saw in the second quarter. And travel sales grew 13%, driven primarily by air sales. Moving to Slide 4. So this is Billed Business Growth by Segment, and we continue to have broad-based growth across all segments. Total Billed business growth decreased about 2%, compared to the second quarter. If we look at USCS, it came in at 12%, down 1% from the second quarter; Global Corporate Products is at 14%, also down 1% from the second quarter; GNS continues to be very strong at 23%, down 2% from the second quarter; and International Consumer came in at 9%, again, 2% down from the second quarter. So we see a slight decline across each segment. But Billed business is holding up well considering the economic uncertainty. Given these strong growth rates, we would expect to gain share again this quarter in the U.S. Moving to Slide 5, this is Billed Business Growth by Region. So Billed business growth by region on an FX-adjusted basis, if we look at Asia, it came in at 19%, the same growth rate as the second quarter. Latin America and Canada is at 14%, down 1% from the second quarter. The U.S. is at 13%, again, down 1% from the second quarter. EMEA came in at 8%, down from 11% in the second quarter. However, we continue to have strong growth in a number of countries in Europe. Germany grew 10%, France 8% and U.K. 12%. Now they're each down between 100 and 300 basis points. On an absolute basis, it is good growth. So we have very good broad-based growth across the regions, although more softness in Europe. If you look at Slide 6, so this is Lending Billed Business compared to Managed Loan Growth. So the solid line is the growth in spending on lending products, and this is at 14% in the third quarter. That compares to 18% in the second quarter of this year and 15% in the third quarter of 2010. The dotted line is growth in loans, and you can see that it is at 2%, the same as the second quarter. We continue to have a spread between the growth in spending on lending products and the growth in loan balances. This is due in part to our actions as we have changed our strategy to focus on Charge Card and premium lending. So just as an aside, Charge Card Billed business and receivable growth continue to move in tandem. But getting back to lending, our lending focus on premium lending and the reduction in balance transfers results in a base that is more inclined to be transactors and revolve less. So if we went back to the second quarter of '08, transactors would have represented about 16% of our portfolio. In the third quarter of 2011, transactors represent 29% of the portfolio. If we were to look at the lending trust and look at payments rates, in September, the payment rate was 31.23%, and that's up from September of 2010 when it was 29.28%. So the combination of our focus on premium lending and the fact that cardmembers are deciding to delever is resulting in the spread in the growth rates between spending and loan balances. For us, the net effect is a lower risk profile. So if you look at Slide 7, which is Net Interest Yield, in the third quarter of this year, it is very much in line with our previous guidance. It comes in at 9%, and we had indicated that yield would migrate back to historic levels either in the high 8s or 9%. They reflect the changes in pricing that we did related to the CARD Act, and also cardmember behavior. Going forward, yield will continue to be influenced by a number of factors, including credit quality, revolve rates and the cost of funds. They are also a subject to regulatory review through the look back. The stability that we see in yield and the 2% growth in loan balance is resulting in net interest income in the second quarter being flat with the second quarter of last year. If we move to Slide 8, Revenue Performance, discount revenue reflects the 16% growth in Billed business growth. But the reason that this is growing at 12% is a result of the slight decline in the discount rate, the relative faster growth of GNS Billed business, and for GNS, we get a part of the discount rate and higher contra-revenue items, including corporate incentives and partner payments. Now the discount rate is at 2.54, the same as the second quarter of this year, but 2 basis points lower than the third quarter of last year, driven primarily by a change in mix. As we have said in the past, the discount rate will decrease as we succeed in our everyday spend strategy and to the extent we have a change in mix by country. Net card fees grew 6%, primarily reflecting a mix shift to higher fee cards and slightly higher proprietary cards-in-force. Travel commissions and fees are down 1%, and this reflects increased worldwide travel sales, offset by a decrease in sales commission rates, as well as the fact that last year we had revenue recognition upon the signing of certain supplier contracts. Other commission fees are driven primarily by the fact that we have acquired Loyalty Partners. So their fees and commissions are in this year, but were not in last year. Other revenues reflect higher GNS partner royalty revenues, higher foreign exchange fees and higher merchant fees. Net interest income is flat, and this is the first time in many quarters that we are flat, and this is a result of the slightly -- the slight growth in loans, offset by a slightly lower yield compared to last year. Total revenues net of interest growth of 9% is notably better than the other issuing competitors and it reflects the difference in our business model where we focused on driving fee income and spend compared to them who are more focused and reliant on net spread income. So fundamentally, we have a different business model. But generally, when we talk about our revenues, we indicated about 80% of our revenues are fee-related and 20% are related to net spread. But in fact, in this quarter, net interest income only represents about 16% of our revenues where our competitors often have between 50% and 80% of their revenues dependent on net interest income. So if we move to Slide 9, Provision For Losses, we can see that the Charge Card provision is at $174 million, up from $84 million last year. That's the result of slightly higher dollar write-offs in this quarter, but it also reflects a lower reserve release that we had last year. So Charge Card improved faster, and most of the benefit from reserve releases have taken place at this point, so the major difference is higher reserve releases in the third quarter of 2010. But credit performance is really excellent and credit metrics are at historic lows. If we look at Cardmember Loans, our lending provision, it is at $48 million compared to $261 million in the third quarter of 2010. So the third quarter of 2011 reflects lower dollar write-offs. So the dollar write-offs in this quarter were $427 million, and that compares with $809 million of write-offs in the third quarter of 2010. Now that is partially offset by us having lower reserve releases in this quarter this year compared to last year. And again we see excellent credit performance and really credit metrics and that are at historic lows. Moving to Slide 10, so this is Charge Card Credit Performance. You can see that the write-off rate in Charge Card portfolios, both the U.S. Consumer business, International Consumer and Global Corporate Products have increased slightly, both sequentially and compared to last year, but they remain at historic lows. So the slight increase in Charge Card metrics align with the company's Charge Card strategic objectives of growing and expanding this business. Our objective is to grow the business profitably and not to have the lowest write-off rate, so you should expect these metrics to increase in the future as we grow this business. Moving to Slide 11, Lending Write-off Rates, so the lending write-off rates continue to improve, both in the U.S. and in the International Consumer business. In the third quarter for 2011, the write-off rate is 2.6%. That compares with 3.1% in the second quarter of this year and 5.1% in the third quarter of 2010. If we look at the lending trust in the U.S., the write-off rate in September was 2.3%, down from 2.7% in August. Again these are historic lows. If we go to Slide 12, we look at lending 30 Day Past Due, the 30-day past due in the United States is stable in the third quarter compared to the second quarter and the International Consumer metric continues to improve. Here, if we were to look at the lending trust and look at 30-days past due, that number in July was 1.5, improved in August to 1.4 and in September was back to 1.5. So I think we are seeing stabilization of this metric in the current quarter. If we move to Slide 13, so this slide is intended to give some insight as we look forward. If we start on the right-hand side, we can see that bankruptcies continue to trend down. And I'd remind you that at the time that we're notified of bankruptcies, generally 2/3 of those accounts have already been written off. So if we look to the left side and look at the upper left chart, and these are balances that roll from current to 30-days past due. If you look at the green triangles, so these are accounts that rolled in February, March and April of this year, these accounts will write off 5 months later if they continue to be delinquent. So the green triangles wrote off in the third quarter of this year. If we look next to that, to the next 3 blue triangles, you can see that they are slightly lower than the green triangles. If the 30-due past day -- 30-day past due to write-off rate, which is the bottom left chart, remains constant, then we would expect in the fourth quarter of this year for write-offs to be slightly lower than the third quarter, but that's also dependent on bankruptcies and recoveries remaining unchanged. So the pace of improvement in our credit metrics have slowed, and we have now reached all-time lows. Again our objective is not to manage our business in order to minimize write-off rates. Going forward, we will continue to balance our risk profile with the company's strategic growth objectives, and over time, we do not expect loss rate to stay at these historic lows. Moving to Slide 14, so this is looking at Lending Reserve Releases. So the chart shows lending reserve releases across the top by quarter. The light blue bar are the dollar write-offs in the quarter and the dark blue bar is provision, which are the write-off dollars less the reserve releases. So the third quarter reserve release was approximately the same as in the second quarter, but less than we saw in the first quarter of this year as the improvement in metrics have started to moderate or stabilize. In the third quarter of 2010, we had $500 million of reserve releases. Going forward, we would expect reserve releases to diminish. Going to Slide 15, so this is Lending Reserve Coverage, and we can see that as credit metrics continue to improve, reserves as a percentage of loans and reserves as a percentage of past due have come down, both for U.S. Card and worldwide. So although we had reserve releases of $400 million in the quarter, coverage remains at appropriate levels. Moving to Slide 16, Expense Performance. Total expenses increased 13% or 11% on an FX-adjusted basis. Marketing and promotion decreased 13% from the third quarter of 2010 when we hit record levels of spending. Marketing is still 10% of revenues, above the historic level of about 9% of revenues. So we still have healthy levels of investment, and I'll talk about this more in a moment. Cardmember rewards and service cost increased 25%, but at a lower growth rate than the second quarter. This reflects rewards-related spending, both in MR and on co-brand products. Plus, within MR, it also reflects the impact of a small increase in the ultimate redemption rate, which reflects higher engagement and higher redemption levels. The cost of cardmember service increased, reflecting the cost of new benefits for U.S. cardmembers. Operating expense increased 15%. But if you exclude the impact of the Visa/MasterCard settlement, this expense, operating expense, grew 9% year-over-year, a notable slowing from the growth rate of 21% in the second quarter. We expect to further slow operating expense growth rate towards the end of this year and into next year, and I'll speak about that more in a few minute -- moments. The tax rate of 28% reflects a benefit related to the distribution of foreign earnings with the associated realization of foreign tax credits. We increased our investment levels in the quarter above previously planned levels in light of this benefit. So if we move to Slide 17, so this is some more detail on marketing expense. And here, each of the bars reflect spending by quarter. As discussed in prior quarters, the high level in investment spending in marketing and promotion has been enabled by the credit provision benefits from improving credit and the Visa/MasterCard settlement payments. As these credit benefits lessen over time and the Visa and MasterCard payment cease, we have the flexibility to move marketing and promotion back towards historical levels or 9% of revenues. Moving to Slide 18, so this is more detail on operating expense. So if we look at salaries and employee benefit, it increased 18%, reflecting a higher level of employees, merit increases, higher benefits, severance costs related to reengineering and higher incentive compensation. If we look at professional services, this decreased 2%, reflecting lower collection cost as you would expect as credit improves and lower consulting fees. The increase in occupancy cost is primarily driven by Loyalty Partners, which is in this year's numbers, but are not in last year's number. We've also broken out the Visa/MasterCard settlement proceeds so you can see those. In total, you can see that we have made progress on our objective of slowing the growth in operating expense. So reported, it's at 15%. Excluding Visa/MasterCard, it grew at 9%, and that compares with 21% in the second quarter of this year. We expect to further slow the operating expense growth rate towards the end of this year, and into next year. Slide 19, this is more detail on operating expense, and this chart is intended to show the areas where we are investing and growing at a faster rate. And there's a difference between those operating expenses which are intended to drive growth and those which are more BAU operating expenses, which are growing at a slower rate. So at the upper left-hand side, you see new business initiatives and they include spending on Serve, Online Mobile, the Certify, Loyalty Partners, LoyaltyEdge and Business insights. You can also see that we're investing in GNS. And if you look back over the last 8 quarters, we have had Billed business growth of over 20%, so those investments are paying off. We continue to invest in sales force across International Consumer, Global Corporate Products and Merchant Services. Variable tech investments generally have been around the average growth rate as they support growth in both our core business and our digital initiatives. And you can see that support functions and Global Services are growing at more modest rates. So operating expense relates both to our ongoing operations and our growth initiatives. Looking at Slide 20, so this has to do with expense flexibility, and here, adjusted expense are all expenses excluding provision divided by revenues. Now here, we recognize that we are currently at elevated levels. Over time, we expect this ratio to migrant back towards historical levels in 2 ways: First, through the growth in revenues; and second, through expense flexibility related to really all items in the P&L. The continued higher investment levels, which were enabled by our tax credit, is why our rate has stayed at 75% in the third quarter of this year. Moving to Slide 21, Capital Ratios, you can see that Tier 1 common ratio remained at 12.3%, the same as the second quarter. So the capital that we generated through net income and employee plans was offset by the distribution we made through repurchases and dividends, while risk-weighted assets stayed relatively level. So the 2.3% is higher than our target for Tier 1 common, and this is due to the fact that we had fewer acquisitions than we planned. As you remember, our guidelines are to retain about 50% of capital generated to support the growth of our balance sheet and acquisitions and to return to shareholders about 50% of capital generated through dividends and share repurchase. In the third quarter, we increased our share repurchase, given the limited amount of balance sheet growth and the fewer acquisitions, but we continued to have strong capital ratios. If we move to Slide 22, this is Total Payout Ratio. So this is a new slide, and this is the percentage of capital generated in each quarter or year that was returned to shareholders. So if we look at 2008, '09, 2010 and the first quarter of 2011, basically, the return of capital to shareholders was through dividends. In the second quarter of this year, we did repurchases of $750 million, and in the third quarter, we increased repurchases to $1.2 billion, given the level of balance sheet growth and acquisitions. Even with 105% payout in the third quarter, our Tier 1 common remain at 12.3%. Fourth quarter repurchases are projected to be $350 million. As part of our request that we made in our CCAR filing to the fed in January, we requested total repurchases of $3.2 billion -- $2.3 billion, thank you, and so therefore, we have about $350 million left of capacity. If we go to the next slide, Slide 23, looking at liquidity, we can see that we continue to hold excess cash and multiple securities to meet the next 12 months of funding maturities. So we're holding $22 billion in cash and securities, and our maturities over the next 12 months are $19 billion. If we slide -- moving to Slide 24, so this is our deposit programs. So total deposits increased by about $1 billion from $31.6 billion to $32.5 billion in the quarter. Since we did not have balance sheet growth, it did not warrant additional or higher funding level. As we've discussed, deposits provide greater diversity in funding for us. So with that, let me conclude with a few final comments. Results for the quarter reflect a continuation of the positive business trends evident during the last several quarters. Spending growth remains strong and we continue to grow faster than most of our large issuing competitors despite difficult prior year comparisons and a challenging economic environment, especially in Europe where growth slowed but is still very healthy at 8% on an FX-adjusted basis. We also saw our average loans grow year-over-year for the second consecutive quarter and lending loss rates continue to improve. Our strong billings growth, coupled with flat net interest income, drove solid revenue growth, which continue to outpace our large issuing competitors, reflecting returns on our investments and the unique nature of our spend-centric business model. Solid revenue growth, improving credit trends and the benefit of lower effective tax rate provided the opportunity to continue to invest in the business at high levels, while also generating strong earnings. These investments are driving current metrics as we deliver high average spending and grow our card base, as well as build capabilities for the future. We feel very good about our recent performance while we acknowledge that the economic environment remains uncertain. We continue to implement our plan to slow the year-over-year growth in our operating expense as we exit this year and into 2012. And in the third quarter, we began to see some progress towards that goal at adjusted operating expense, which excludes the MasterCard receipt, grew 9% versus 21% in the second quarter. Our success in a highly competitive environment and our focus on spending as opposed to spread are yielding high-quality results and solid revenue growth. Going forward, we will closely monitor our spending and credit trends, but to date, we are very pleased with our results and we are confident that our investments and business model are appropriately positioned to navigate through this environment. Thanks for listening and we are now ready to take questions.

[Operator Instructions] We now go to our first question, Mr. Don Fandetti with Citigroup.

Dan, I was wondering if you could provide a little color on where Billed business trends are in October overall, and then maybe whether or not Europe's continued to sort of soften a bit or if that stabilized.

So spend growth continues to hold up well in the fourth quarter. However, we are just a few weeks into the quarter and we will watch spending trends closely as we go forward in the next coming months.

And on Europe?

I think -- I'd say we're holding up well, although we saw greater softness in Europe in the third quarter. So we'll have to see how that pans out as we go through this quarter.

Next, we go to the line of Bob Napoli with William Blair. Robert P. Napoli - William Blair & Company L.L.C., Research Division: Question on Rewards. The Rewards expense as a percentage of discount revenue was 37% this quarter. And if you go back a couple of years ago, it was running at the 30% level. And, I mean, you've been bringing up your redemption rate. You said you brought it up a little bit more this quarter. And I was redeeming some points myself the other day, it's awfully easy to do. And so I don't know why I wouldn't redeem all of them easily over time, but is there -- how much was the one-time hit this quarter for higher redemption rate? And, I mean, is this run rate, is that permanent?

So we're pleased you redeemed your points. We believe that to the extent cardmembers redeem points, we have greater engagement. They spend more with us. We have lower attrition. So the economics associated with that, we think, are very positive for the franchise over time. So in any given period, the Membership Rewards expense is going to be driven by 2 primary -- well, 3 primary factors. One, where the volumes are. And so to the extent we have volume growth, we're going to have growth in expense related to that. And the other 2 factors are the ultimate redemption rate and the weighted average cost per point. Now to the extent that we are having higher levels of engagement, higher redemptions by customers, we take that new information and put it into our calculation of what we think the ultimate redemption rate is going to be. And to the extent that it's higher, then we are going to have higher expense related to that. I went through a explanation on the call last quarter that says expense is really a combination of the activity and the current period, as well as repricing our bank which is very large, which is a good asset for us, but that repricing results in additional expense. To the extent the redemption rates are going up, we would see growth rates on that expense line in excess of the spend growth in the quarter. So what happens in the future is going to be very dependent on the behavior of our customers in future quarters. If, in fact, it runs at a slightly higher rate, then the growth in Billed business or spending, again, I would view that as a positive. The growth rate in this quarter is less than we saw in the second quarter, and that's because the growth in the redemption rate in the third quarter was somewhat less than the growth rate and the ultimate redemption rate that we saw in the second quarter. Robert P. Napoli - William Blair & Company L.L.C., Research Division: Okay. Follow-up question just on share repurchases, when and how do you go back to request more -- essentially, you bought back 2% of the company this quarter and your capital ratios didn't change. And I know you're looking at acquisitions, so maybe thoughts around the -- I mean, a little color on how you expect to reload the share repurchase authorization from the government, if you will, or from the fed.

Okay, last year, the fed had ourselves and I think all other bank holding companies submit filings, which reflected what our base plan was for the coming year. Also required us to have projections in an adverse and a more stressful environment that they reviewed. We made their request for share buybacks as part of that filing. Now our request for share buybacks was based on this notion that we would distribute about 50% of generated capital back to shareholders and retain about 50% for growth in the balance sheet and acquisitions. So the 2.3 was our amount. Now we performed well during this year. Balance sheet growth has not been significant and we've made some acquisitions, but not at the level that would warrant maintaining 50% of the capital. Now we could have ventured to make a new request to the fed in the recent month or so, but we thought it was more prudent just to address what we think our needs will be next year in the filing that we will do this January related to the required filing, not related to 2012. So we will be mindful of this year's experience as we make our request for next year.

Next, we go to the line of Ryan Nash with Goldman Sachs.

On the expense line, you guys came in at 75%, I think a little higher than some of us have been looking for, but you did mention there was, too, the tax credit. And I know that your guidance calls for -- the rate of change is slow. So can you just help us understand your ability to generate positive operating leverage? And at what point will we see revenue growth exceed expense growth? And I have another one after that.

As I think, we have been in a situation, really, over the past 2 years through 2010, 2011, where we have had very good credit performance. That results, as you know, in the release reserves, and we've had the Visa/MasterCard proceeds. And we've decided to use that money to, in part, improve earnings and, in part, invest in the future to drive business momentum. And that's what's really resulted in that ratio staying up at that 75% level or so. As we slow the growth in expense, as you would expect to see, that ratio come back down. It's going to come back down really in 2 ways. One is the growth in revenues and second is the slowing of the growth in expenses. In this period, if we hadn't had the tax benefit, then we would have ratcheted back investments and you would have actually seen that percentage come back down. But since we knew we had the tax benefit coming, we decided to invest it, and therefore, it starts -- it stayed at higher level. So you would think over the coming quarters, you would start to see that percentage to come down if we execute against our plans.

Okay. And just on the lending side, I think you had alluded to when you've had the conference last month that you would expect that losses would be lower than the 4.5% that you had historically had. But I guess numbers are running significantly lower than that, and your lending strategy has changed so much. So can you just help us understand either in numerically or contextually how you're thinking about where losses will eventually stabilize?

I'll do it contextually.

I figured I'd at least try.

So 4.5% is the rate we had over a 10-year period going back. As you said, we have changed our strategy. We are focused on premium lending. So I would expect it to, as I said before, be below that 4.5%, and I would expect it to be above where we are now. And these are really historic lows. And our desire at the end of the day is not driven by any target for a write-off rate. It is to make good, economic decisions that will benefit the franchise over the long term. That's the way we've always mirrored investment decisions. That's the way we'll continue to make them, and I suspect write-off rates are going to, obviously, settle somewhere between where we are now and that 4.5%. But to actually peg an exact number will be very dependent on the businesses decisions we're making. And as you know, there's a cyclical impact here, and in more robust times this is a lower number. And in times of economic stress, this is a higher number, but I think it will obviously be below where we were historically.

Next, we go to the line of Craig Maurer with CLSA. Craig J. Maurer - Credit Agricole Securities (USA) Inc., Research Division: Two questions. The first, if you could -- as a follow-up to the last question, be a little bit more specific on, did you reinvest the entire tax credit, one; and two, as it relates to that, a little bit more specific on what you might have spent it on so we can see the one-time nature in that spending. Secondly, have you seen any significant impact from the loss of Continental that you can point to?

So we did reinvest the 100% of the tax credit. It wouldn't be one isolated item. We knew that we had that resource going into the quarter. So as we've decided on our investment allocation, we took it into consideration. But you can see in marketing, while it's down from the third quarter of last year, it is at elevated levels. And as you look at expenses on the chart, maybe in Slide 19, you can see that we're investing in sales force, in GNS, in new business initiatives. So it really was part of the allocation of resources across all of those as opposed to just one isolated item. As it relates to Continental, at the end of the day, the removal of Continental from our MR program caused some increase in point redemption. However, quite frankly, the impact of Continental leaving so far has been less than we would have anticipated. I point out that even with Continental leaving the program, our Platinum Card base in the U.S. in the third quarter grew. And we believe that our MR program remains the industry-leading option for high-spending customers. So there was some customers that moved, but less than we expected. And we continue to have a very strong card base of affluent customers who spend at high levels.

Next, we go to the line of Brian Foran with Nomura.

I had a question, but actually I'll ask a follow-up to that. I mean, if U.S. Bank lost Northwest, held together much better than everyone have thought, didn't lose that many customers, you lost Continental, you're saying you're holding -- the cards holding together, you're not losing any customers. Does that imply the industry is just paying too much for these airline partnerships?

Well, I think, as we know, the cost of rewards related to co-brands are higher than our cost of our Membership Rewards program. We think our Membership Rewards program functions very well and is a program that many customers want because it provides a lot of options. Instead of being tied to just one redemption option, there are many redemption options, I think over 100 redemption options, which customers find to be very valuable. But when you look at co-brands, despite the higher cost from a rewards perspective of co-brands, those customers tend to spend at a higher level and the credit losses on those portfolios tend to be at the low end of the range. So the economics associated with co-brand products are good, and that's why we continue to have those programs and endeavor to make them more attractive. So while it was a different rewards mix, at the end of the day, when you look at the economics of a product, you need to look at a combination of what does it cost you to acquire the customer, what are their spending levels, what are their revolve levels, what will ultimately be the write-off rates within that group, determine the overall economics. So while rewards costs are higher, the economics of those programs are very good.

And then just in terms of pricing on cards, the environment is worrying to a lot of people in terms of mail volumes increasing, balance transfer options are out there. I guess looking at this quarter's data, your yields are up 30 bps, USPs are up 20, JPMorgans are up 2, some of the other guys are down, but even in the industry average is flat. I guess any more color you can provide on why the yields were up, 30 bps linked quarter? Is there seasonality in there? Is it the higher current 30 roll rates or is pricing competition not as bad this year?

I think we had indicated that we thought that yields would eventually migrate back to where we were before the CARD Act. And I think we've actually seen that now, but as you know there are a myriad of factors that come into play. Credit quality, to the extent you have more people in delinquency, unhealthy pricing that could drive it up to the extent you have more transactors that can drive it down. Most of our cards are variably priced. So when cost of funds go up or down, it doesn't have a huge impact. But as I said before, transactors are a decent part of our lending portfolio. So their interest costs really do matter. For us, as I indicated, the pay down rate in the third quarter this year improved compared to the third quarter of last year. But in fact, the pay down rate dropped out a little bit, 20 to 30 basis points, from the second quarter. So the fact that the revolve rate changed a little bit, I think that had an impact on the higher yield that we saw. But I think it's going to move around quarter by quarter based on all those factors. But to the extent it kind of stays in that range of the high 8s to 9%, I think that enables us to have good economics related to those products. And so I wouldn't focus so much on the little moves quarter to quarter, but more that we can sustain the appropriate yield over time in terms of assessing the economics of those products. I guess the other thing that I would note is that while others are doing BTs, we are doing very, very few in that area because our focus is not bringing on customers who are going to sit with a loan balance. Our objective is to bring customers who spend and occasionally revolve. And that is really at the heart of the difference between our business model and a competitor's business model. We want to generate fees from spending and allow our customers to revolve if they choose to, but it's not to bring customers who are going to sit with balances. We think that will put us in good stead over the long term because we think competitors, quite frankly, have a challenge. If you look back in history before the recession, loan balances were only growing at about 5%. And we can certainly see within our portfolio that there has been a separation between the growth in spending and lending products and the growth in loans as customer or consumers have decided to deleverage more. So that's going to present the challenge for them. But I think our business model is right for the environment as we go forward.

Next, we go to the line of Betsy Graseck with Morgan Stanley.

A couple of questions. Just one other angle on the expense question is that you are suggesting that you will be bringing it down this quarter. Can you just give us a sense of where you think you can do that in a way that doesn't impact your revenue growth, in particular, during this heavy spend holiday season. I'm just wondering where the flexibility is.

All right. Okay, so we haven't said that we're going to bring spending down. What we've said is that we are going to slow the growth rate in operating expense growth. Okay, so there's a little bit of distinction there. But I do think that we do see some opportunities. I mean, I think, we saw it in the third quarter, right? So in the third quarter, we saw that as credit is better, collections cost came down. I think we have brought down consulting services. So that's something that's very discretionary. We can decide to spend that at a higher level if we think we have an investment that we want to make in a new area or to enhance a product. But that's very discretionary at the end of the day. We also continue to be at elevated levels of spending on marketing and promotion. In this quarter, it's about 10% of revenues. Historically, we've been able to generate business momentum with marketing and promotions being about 9% of revenues. And then, as we've discussed, we set up Global Services group, which is really intended to be very focused on high-quality service by continually improving the efficiency in the way that we do things. We've set targets for that group that history has spoken about. We are being very successful at executing against those targets and achieving those goals that we've set for ourselves. So it's really across each of those types of areas that would enable us to slow the growth of operating expense while still having appropriate levels of resources to continue to drive the growth and the business momentum that we've generated in 2010 and 2011.

Next, we go to the line of Chris Brendler with Stifel, Nicolaus. Christopher Brendler - Stifel, Nicolaus & Co., Inc., Research Division: The recent hoopla on the debit card fees, that this may be causing some disruption in the debit card market. I remember, a year ago, you started to position your Charge Card as been tried as an alternative [ph] to the debit card. And I wasn't sure if you had any update on the results of that campaign and when you were thinking that this could be a potential opportunity for American Express to be able to share given what's going on in the debit card market.

I think we've also talked about the fact that Charge Card has some features that are very similar to debit and there are some aspects that are different. So while it's a pay-in-full product, it's not immediate, but there are some similarities. And if you're sufficiently disciplined to pay in full, then you can also get reward points and other services that are not available if you utilize the debit product. So I think a key here is that we've never been in the debit business. The profit pool in debit was never that large. With the changes required by Durbin, whatever profit pool was there is gone. So banks are looking to say, "How do we look at the overall economics of our relationship with a customer that they have?" And to the extent that relationship is relatively limited, then I think they've discussed that they really don't have the ability to provide that service for free without some kind of a cost. So I think what we're going to see is, over time, potentially a effort on the banks of -- on the part of banks to either charge for that service or to migrate customers either to Prepaid. And as you know, we have a very strong Prepaid offering. We issued a new Prepaid product that if you acquire that Prepaid product online, there are no fees to acquire it and no fees over time. So if people move in that direction, we have a product that is available. To the extent that banks push customers more towards credit, we think that, that would be an opportunity for us to grow our business as well. So we have to wait and see how this all plays out, but it could well be that this turns into more of an opportunity than a risk at the end of the day, although the whole script hasn't been written, so we'll have to see how that plays out. Christopher Brendler - Stifel, Nicolaus & Co., Inc., Research Division: It's fair to say that at this point there's no plans to restart that campaign. I guess I was also wondering were you happy with that campaign and the results you got? It seems like it is a different product from a consumer standpoint and I wasn't sure if the message you were sending last year really translated to consumers. A follow-up question would be actually on Serve. When should we expect Serve, if you had any update on it, when is it -- the volume metrics or any disclosure around Serve and how that's going -- that it starts?

Okay. So I don't think we ever really had a campaign to say that the Charge Card is the equivalent of a debit product. I think in different conversations with investors, we probably talked about the fact that there are some aspects of the product that are similar, but the fact that it's a payment in full when you receive the statement, I would say it's more fundamentally a different product. And the Charge Card product is maybe really targeted at different set of consumers than consumers who are using the debit product. Debit products, I think, are probably more for smaller ticket item, for people who have more limited resources and are trying to manage them very carefully. As you know, our Charge Card products are really targeted at more affluent customers. They have fees attached to them, but they also have substantial benefits associated with them. And we have, on the other hand, had a campaign to grow Charge Card over the last year and a half, and we have been really very successful at acquiring high-spending cardmembers and growing cards. And you could also see the Charge Card spend in this quarter grew by 14% compared to the prior period. So we're being successful at acquiring cards and having our customers grow their spending on those cards. So in terms of Serve, your second question, what we have been targeting over this year, 2011, is to, first, launch Serve, which we did in March. You remember that we acquired Revolution Money in January of 2010. From early 2010 through March of this year, we looked at building on the platform, in building the capabilities. The launch that we did in March is really kind of Serve 1.0. We're going to look to enhance that platform and we've been focused on building the capabilities on that platform throughout 2011. The metrics that we put out there that we wanted to accomplish in 2011 was really to enter into a number of partnerships where we can actually drive customers in the future to the Serve platform. So this year, we announced the partnership with Verizon that integrates Serve into their pay-by-phone service and they're embedding Serve in new Verizon phones. We also announced the partnership with Sprint to embed Serve into the Sprint Zone. So we've signed also some agreements with AOL as well. So we are kind of hitting, in our mind, our objectives for this year to improve the platform and to sign new partnerships that will enable us to bring customers on to our network and to have Serve product. So I think for this year, those are the guideposts that I would look to for success. As we look to 2012, I think, then we're going to look to actually seeing how successful we are actually bringing customers onto a Serve product and onto our network. So those are the guideposts that we are looking at in terms of success.

Next, we go to the line of Sanjay Sakhrani with KBW. Sanjay Sakhrani - Keefe, Bruyette, & Woods, Inc., Research Division: Maybe I could just follow up on that rewards cost question asked previously. Despite my rough math, I think the redemption rate moved up to about 93%, and I guess I was just wondering like how much higher can it go given the current trends? And secondarily, just to the extent that you do need to take that redemption rate assumption up, you guys still feel comfortable that you could hit your historical kind of efficiency ratio targets as shown on Slide 20, in that scenario. And then maybe just a second question to that is just reserve coverage. And obviously, you had very strong coverage levels right now. What is the normal level if we head into next year?

Okay. So on the first question, you remember that in last quarter, we indicated that we were at 92% related to active customers. And that in the first quarter, we actually rounded up to 92% and we said we had growth in the ultimate redemption rate and that we kind of round it down to 92%. So in this quarter, we again had growth in the ultimate redemption rate, but we're still rounding down to 92%. So it grew, but still rounds down to 92%. So we're not at 93% yet. Now how high can it go? That's difficult to forecast. It will be dependent on the behaviors of customers in the future. So certainly, customers who have a trend [ph from their program already, on average have tried it at a rate that's below 92%. But we look at the behaviors of the active customers. And then we draw our curves into the future in terms of what we think the ultimate redemption rate will be. So the 92% represents those behaviors to date. As customers become more engaged, it will potentially drive that rate up. Will it ever get to 100%? I think that's probably unlikely. But based on customer behaviors, it certainly could go above the rates that we are at today. So the question about how does it fit into the efficiency ratio, so we can look at efficiency ratio, but I think we need to look at really all expenses, including provision. So I think that we do need marketing and promotion over time on average, and over time to kind of be 9% of revenues. On the other hand, because write-off rates are going to be lower than historic levels, then I think that's going to allow the efficiency ratio to probably be a little higher than it may have been prerecession. But we're going to need to balance how we decide to utilize our resources between Rewards and operating expense over time. And that's a judgment we'll make in the normal course of business in terms of how we decide what to allocate to marketing, what to allocate to Rewards, what to allocate to other business initiatives to build the capabilities. And that's a map and planning that we are doing now, have always done and will do in the future. And we believe that even with higher Rewards cost, that we can continue to achieve our current financial objectives of revenue growth of 8% plus and 12% to 15% EPS growth and an ROE of 25% plus. So even with the high Rewards cost, we'll be -- -- had planned stay with our current financial targets. So I think... Sanjay Sakhrani - Keefe, Bruyette, & Woods, Inc., Research Division: And the reserve coverage?

Okay, reserve coverage. I think reserve coverage is obviously lower than it was a year ago or earlier this year. As credit improves, I would expect those ratios to gradually come down. This is an uncertain environment and so we're cautious as we release reserves. But to the extent we get to a point where there's less uncertainty, then I think you could potentially see lower reserve coverage than we have at the moment, but that will be driven by the fact and circumstances at the time.

Next, we go to Ken Bruce with Bank of America.

I was hoping you could provide a little bit more detail or decompose the corporate incentive payments and partnership payments that was in the contra revenues item this quarter. If there was one-time items, size it up in any way for me?

So I don't think there were any one-time items. I think we interact with our corporate customers and particularly our larger corporate customers, and we provide them incentives to do business with us. And certainly to the extent that their volumes increase above certain threshold levels, we provide incentives related to that as well. So these incentives are all around driving higher spend levels at the end of the day. So each quarter, we have new signings. And those new signings are reflected in the results. But again, even with these higher incentives that we're paying to corporate customers and partners, the economics around these agreements are very good. And we make that evaluation in terms of the returns that we're achieving and ensuring that they achieve our return thresholds. So I don't have more specifics than that, but these are good economic decisions that enable us to both grow the business and achieve our financial targets that we just discussed a moment ago. So I would think about these probably not differently than I think about Rewards. At the end of the day, you want to grow volumes and achieve the right economic returns, and these agreements are enabling us to do that. So let me just have this be the last question.

Okay. Can I have a follow-up?

Okay, go ahead. Well, one follow-up and then one last question.

Okay. And it will be a 2-part follow-up. On the discount rate, you've, in the past, guided to a few basis point erosion per year. I think as you pointed out in your earlier comments, much of that was a mix change issue on a year-over-year basis in the quarter. I'm wondering is there an order of magnitude that you could provide in terms of the net discount rate which would include the contra revenue items as we think about forecasting? And then as a second part, how do you think about the overall level of transactors in your overall lending book? I think you pointed out it had grown from 16% to 29%. Based on the number or the percentage of mail drops, where do you think that ultimately gets to?

Okay. So on the first question, in terms of the discount rate, to the extent we are successful and continuing to drive into new categories or are able to create greater engagement of existing customers into new categories, they will, in all likelihood, drive the average discount rate down by 2 or 3 basis points a year. Over the last 3 or 4 years, it's only been about 1 basis point as we've been successful or we're historically successful at increasing price in certain categories where it was warranted based on the value that we bring. There are less of those opportunities at the moment. And that's why I think we're seeing the 2% lower discount rate -- 2 basis points lower discount rate. But again, the economics are good here. It's part of our strategy. This is something that we are driving as opposed to just happening to us at the end of the day. I actually have not really thought about where it might go net of incentives. It's not something that we have done. So that's not a statistics that I thought about quite frankly. The other thing is that as you do the calculation where you're just taking discount revenue divided by billings, you need to be mindful of GNS. So GNS is a terrific success story. It is driving additional business. It's increasing relevance of our network around the world. But as you know, in GNS, we get a percentage of the discount rates and not the full 2.5. That's impacting that calculation. It's probably the biggest impact on that calculation. But the economics of the GNS business are very good. As you know, we don't have credit risk and has very good returns on equity. So we shouldn't think about that ratio going down as a negative. It's actually, again, driven by our business strategy. And we are being successful. In GNS, as you know, we've had growth in Billed business of over 20% for the last 8 quarters. Now in terms of transactors, you're right. Back in the second quarter of '08, transactors only represented 16% and that is now up to 29%. I think that is a combination of our strategy, which we are focused on premium lending. And I think that would cause that number to go up over time. On the other hand, another very big factor is just customer behavior where customers have decided to delever. Now whether customers continue to be in that load over the long-term or whether over time, they are more willing to take on additional revolving credit. That will drive that number as well. So I think it will be both of those factors over time that drive where that percentages go. Again the important thing is that we have good economics, whether it's a customer who is spending and revolving or a customer who is simply a transactor. So now, I'll take one last question. Okay? So if we have no further questions, I thank everybody for being on the call. Thank you very much.

Okay, that's a wrap.

Ladies and gentlemen, this conference will be available for replay after 7:00 p.m. this evening through October 26 at midnight. You may access the AT&T Executive Replay System at any time by dialing 1 (800) 475-6701 and entering the access code 218010. International participants, dial (320) 365-3844. That does conclude our conference for today. Thank you for your participation and for using AT&T Executive Teleconference Service. You may now disconnect.