PRA Group Inc (PRAA) 2017 Q2 法說會逐字稿

Good afternoon, and welcome to the PRA Group Earnings Conference Call. (Operator Instructions) As a reminder, this conference call is being recorded. I would now like to turn the conference over to Ms. Darby Schoenfeld, Vice President of Investor Relations for PRA Group.

Thank you. Good afternoon, everyone, and thank you for joining us. With me today are Kevin Stevenson President and Chief Operating Officer; and Pete Graham, Executive Vice President and CFO.

We will make forward-looking statements during the call, which are based on management's current expectations. We caution listeners that these forward-looking statements are subject to risks, uncertainties and assumptions that could cause actual results to differ materially from our expectations. Please refer to the earnings press release and our SEC filings for a detailed discussion of these factors. The earnings release and the slide presentation that we will use during today's webcast and call and our SEC filings can be found on the Investor Relations section of our website at www.pragroup.com.

Additionally, a replay of this call will be available shortly after its conclusion, and the information needed to listen is in the earnings press release. All comparisons mentioned today will be between Q2 2017 and Q2 2016 unless otherwise noted. I'd now like to turn the call over to Kevin Stevenson, our President and Chief Executive Officer.

Thank you, Darby. And good afternoon, everyone. Thank you for joining us on our 2017 second quarter earnings conference call.

Before we discuss the results of the quarter, there are a few items I'd like to mention. First, I cannot begin my tenure as CEO without publicly acknowledging the incredible job that Steve Fredrickson has done over these past 21 years. Founding PRA was Steve's idea, and none of us would be on this call this evening if it were not for his vision, hard work and determination. I've worked side by side with Steve for over 2 decades. When I look back on those years, I realize how fortunate I've been to have joined him on this journey.

Steve has now moved into the Executive Chairman role, and I want to take just a moment to thank him for his many years of leadership, vision and hard work. Steve commented last quarter that it was his 58th and final earnings conference call as CEO. Today, marks my 59th earnings conference call, and my first as CEO. I thought it appropriate at this time to reiterate just a few of PRA's long-held philosophies.

First, we do not name the company who we purchase from, nor will we identify companies that are generally in or out of the market. We've held this philosophy since our IPO in 2002 for a host of reasons, one of which is that selling distressed assets is a strategic decision for the financial institutions, and any color on that decision should come from them not us. Second, in our founding in 1996, we made a fundamental decision to take a long-term view of this industry, to create a company that would compete in the right way, for the right reasons and do it for the long term. We believe this perspective has allowed us to avoid any number of perilous actions over the many years we've been in business. Third and finally, when we listed in 2002, we made an active decisions -- decision to not provide earnings guidance. Instead, we offered to produce data for the investors to digest and analyze.

We recognized then and now that not giving earnings guidance has its pros and cons. But we continue to believe it's the right philosophy, allowing us to focus on making the right decision for the company and giving investors the data to assess the business.

Related to this, we've taken steps in the past few quarters to help investors better understand the revenue accounting model. While we've done this in the past, it seemed like a good time for an accounting refresher. We included a slide in the June William Blair Conference Presentation that walked investors through how our financial receivable revenue is computed under GAAP.

While the slides from the Blair Conference is more accounting focused, the slide now on your screen shows how the revenue calculation is performed using the accounting guidance and the information available on our filings. Our goal here was to demonstrate how Q4 and Q1 data could be used to predict Q2 results. We used our Q4 2016 and Q1 2017 supplemental data tables from our 10-Q and 10-K in preparing the slide.

The math produces $183.6 million as a base revenue estimate for Q2 2017 and FR revenue. To build on that estimate, you must make a few assumptions. And these are quarterly buying, quarterly yield changes, cash from fully amortized pools and allowance charges or reversals. In the slide on your screen, we simply used Q1 2017 results as a proxy for those estimates. As you can see in our exercise, this method computed NFR revenue at $188.2 million or within about 1.4% of the actual result of $190.8 million. We thought this demonstration might be helpful because it's a way to check your existing models for reasonableness for the upcoming quarter. The complete calculation is in the appendix to these slides.

Now on to the results of the quarter. Portfolio purchases in the quarter were significant. U.S. insolvency had its second straight quarter of excellent investment volume. Looking back upon Q1 2017, our $67 million investment represented the largest amount deployed since Q1 of 2014. Now in Q2, we surpassed that amount by nearly 50%, for a total Q2 investment of $100 million. This is our largest insolvency investment quarter in PRA's history by about 15%, excluding an M&A transaction.

We are very pleased of this outcome. However, we believe it's important to understand the circumstances. The increased bankruptcy buying in Q1 and Q2 resulted from large portfolio purchases from one seller. We believe our size, experience and relationship helped us prevail on these purchases and, of course, a sharp pencil. I want to thank everyone in our data and insolvency department that worked on these acquisitions over the last 2 quarters. I sincerely appreciate your efforts.

Core purchasing in the Americas was $145 million, and just like Americas insolvency, it was the largest single quarter in company's history. We are very pleased with this result, but unlike insolvency, this performance was driven by general market conditions in the United States.

The volume increases we are seeing in Americas core are coming from current sellers in the market and not from the return of any sellers. We have no further insight into when any of the sideline sellers may plan to return. Additionally, we are aware of recent commentary that a seller is exploring, taking post-charged-up collection in-house, but that does not appear to be an overall market trend.

Overall, we are encouraged by what we're seeing in volumes. And as we discussed last quarter, this appears to be a natural progression of the seasoning of the lenders credit card loan portfolio. We currently see nothing unusual from a macroeconomic or consumer perspective impacting credit cards.

As we've been saying for the past few quarters, underwritten returns on new portfolios have been steadily improving versus last year, and the competition in the U.S. remains steady and rational. This is very important, as irrational players have historically plagued this industry prior to the current regulatory environment. This is an asset class that we believe deserves and demands a long-term view, and participants need to bring more than just a checkbook to the table. We are not trading portfolios of generic securities. These are individual consumer accounts that require attention, understanding and effective processing. I hope that federal and state regulators as well as sellers understand our position on this matter. We are willing participants in this increased regulatory environment and focused on compliance and the customer's journey to recovery.

On the operations front, starting with the Americas. The insolvency cash headwind is lessening. In 2015 and 2016, America's insolvency cash collections declined on average $26 million quarter-over-quarter. The past 2 quarters this has decreased to $19 million and then $15 million this quarter, as we begin to anniversary the exceptional years of buying we experienced in 2009 to 2013. Additionally, we are pleased the cash collections increased sequentially in Americas insolvency for the first time since Q2 2014.

Insolvency operations remain solid and highly scalable, allowing us to both service the record buying of the past 2 quarters and prepare for more. Americas core operations continues to increase its year-over-year results. All indications are about the significant recent hiring of collectors is doing what we had planned, as cash collections increased this quarter in our U.S. call centers by more than $3 million, reversing the trend of the last 4 quarters. Our U.S. collector headcount is now over 2,200, and our sites are at capacity.

Given the significant buying with experience coupled with our forecasted buying in the near future, we are working on opening 2 new call centers, the first as early as Halloween. As expected, with our headcount build, productivity is down and more in line with 2014 and 2015, as we continue to hire. Additionally, recall from our prior conference calls that we purchased a record number of accounts over the last 3 calendar years, averaging almost $3 million a year in the U.S. These accounts have also had lower average balances, making them more suitable to collect in call center versus legal, thus shifting even more volume towards the call center.

This year, we are on track to meet or exceed $3 million accounts purchased again. On the regulatory front, we are waiting for a decision from the D.C.Circuit on the ACA versus FCC lawsuit, regarding the Telephone Consumer Protection Act or TCPA. Apart from that decision, the commissioners of the FCC appear to understand the business need in regard to the TCPA, and we remain hopeful we will be able to use technology in our collection efforts.

The CFPB has indicated it will issue its notice of proposed rules in regards to -- in regard to the collection industry possibly as early as September. The process then consists of a comment period, writing of final rules and likely an implementation period. We are unsure of the timing for final rules but will follow the process carefully.

Finally, before I move to Europe, just a quick comment on the IRS. We're very happy to put this significant piece of litigation behind us. The outcome is acceptable to us, and we believe it's in the best interest of our shareholders.

Moving on to Europe. Europe remains competitive, as many push into new geographies to create buying activity. We are starting to see sellers enter multiple year for flow agreements, which tells us they think pricing is high. Additionally, contract terms are increasingly onerous. We will continue to focus on bidding strongly on ideals that are justified at market pricing. This will likely result in less buying than we had originally expected in Europe in the near term, but we plan to be in this market for the long term and provide sellers with a reliable, trusted, consumer-focused partner.

In Italy, we continue to make progress on legal collections processing. However, it is low due to the complex nature of the legal system. The pools we placed on nonaccrual in early 2016 are still on nonaccrual. Recall, that in Q4 2014 and Q2 2015 we invested approximately EUR 100 million in Italy on 2 large portfolios. We've collected EUR 50 million on these portfolios since we acquired them.

SME portfolios are also on nonaccrual across Europe. SME is an asset class that holds a lot of market opportunity, but it is new to us, and we'll take time to develop the data and ability to reliably participate in this market. This is similar to what we did with our U.S. bankruptcy asset class in 2003 through 2004.

Now I'd like to turn the call over to Pete to go through financial results.

Thanks, Kevin. Cash collections as a whole were generally in line with expectations, with increases from Americas and Europe core and Europe insolvency on a currency adjusted basis. Total cash collections for the quarter were $375 million, a decrease of $13 million compared to the prior year but only $7 million when adjusted for currency. The decrease was driven by a $12 million decline in global insolvency cash collections, as we continue to move away from the 2010 to 2013 time frame where we purchased significant amounts of U.S. insolvency accounts.

This has been a constant headwind for us, and we're very encouraged by the sequential increase in U.S. insolvency cash collections from the first quarter, something that hasn't happened in 3 years. While the cash collection decline is lessening, remember that the portfolios that are running off were higher yielding, so the revenue numbers may not move on a one-to-one basis with the cash.

Americas core collections were up $3 million versus the second quarter of 2016 at $217 million, driven by growth in Brazil and U.S. call centers. The impact of collector hiring was reflected in the increase in U.S. call center collections of $3 million. We've been purchasing a record number of accounts with lower average balances and have improved the legal selection methodology. This is decreasing the overall number of accounts that are being placed in the legal collections channel, resulting in legal collections being $5 million lower than the prior year.

The trends we're seeing in cash collections in the call centers tells us that our staffing increases are starting to have the desired effect of increasing collections and building payment plans. Our focus is to ensure we have sufficient capacity to provide for our current portfolio and expected growth.

Keep in mind that because we've had significant buying in U.S. core in the first half of 2017 our call center capacity has been stretched, which is why we're planning to open new call centers. We considered the impact of record numbers of accounts being purchased and have allowed for additional ramp-up time when booking these deals. Because of this, we anticipate the cash in revenue to be a little more delayed than we would traditionally expect. Europe core collections were $99 million, down $4 million from the previous year but up $2 million on a currency-adjusted basis. We saw variability country to country and continue to face a very competitive environment on the purchasing front. We still have the pools in Italy and those that contain high concentrations of SME on nonaccrual.

Operating expenses were $152 million, down $4 million from the previous year. The decrease was mainly driven by legal collection expenses and agency fees. With fewer accounts being placed into legal collections this year and our improvements to the legal selection methodology, our U.S. legal collection expenses were lower. This was partially offset by increased legal collection expenses in Europe, specifically Italy, where we had lower agency fees because we moved more accounts into the legal channel.

At this point, we aren't providing any specific forward-looking metrics on legal collection expenses, only the points mentioned here. The decrease in legal collection and agency was partially offset by an increase in compensation expense, resulting from hiring in the U.S. and increased outside fees and services. Included in outside fees and services were several nonrecurring expenses. In pretax dollars, these were $2.3 million in legal fees not associated with normal operations; $1 million in expenses associated with PRA location services, which we sold in June; and $700,000 in financing cost expensed in the quarter.

Our cash efficiency ratio was 60.1% in the quarter versus 61% for the full year of 2016. This is very positive considering all the additional people we've hired. Over the short term, we expect additional pressure on this metric, as we continue to hire more collectors and open new sites and process the record volume of insolvency accounts we've recently purchased, which have a front-loaded cost profile.

One other item I'd like to mention is the $1.9 million tax impact that occurred in the quarter. As the mix of our income shifts more towards the U.S., we adjusted our year-to-date provision for income taxes to reflect that change in the earnings mix. Additionally, we settled the IRS lawsuit. And I want to remind you that this primarily relates to the timing of tax payments and that barring any change in tax rates it will not impact the amount of taxes owed. The impact of the deferred revenue being brought into income will be spread over 4 years, so we don't believe it will have an impact on our ability to purchase portfolios.

Estimated remaining collections totaled $5.3 billion at quarter-end. You can see the breakout on the slide. We have 55% in the U.S. and 42% in Europe, the remainder being other Americans. ERC increased almost $200 million sequentially and like many other metrics was declining in the second half of 2016, hit an inflection point in the first quarter and is now moving positively.

After the capital raise we completed during the quarter, we have dry powder of $600 million in the Americas and $538 million in Europe available for portfolio purchases, a total of $1.1 billion worldwide. We stand ready and committed to helping sellers manage their charged-off debt.

Before I hand things back to Kevin for a few closing comments, I wanted to return to the revenue model we provided earlier. The 10-Q will be filed overnight, so you'll be able to do this calculation for yourself but thought I would walk you through the details one more time. While this slide is busy, it shows you a base estimate for third quarter NFR revenue of $187.1 million. Remember, there are 4 items that can impact this number: revenue from portfolios purchased during the quarter, cash collected on fully amortized vintages, allowance charges or reversals and yield raises.

Now I'd like to turn things back to Kevin, for a few closing remarks.

Thank you, Pete. As the first half of 2017 comes to an end, I want to remind everyone that our business has transformed over the last 3 years. It began with guidelines from the OCC on debt sales in 2014 and the beginning of the rule-making process from the CFPB. We faced supply headwinds, as many of the banks evaluated both of these and experienced historically low charge-off rates. Later in 2014, we acquired Aktiv Kapital and significantly increased our business overnight.

Shortly thereafter, in late 2015, the requirement for more comprehensive consumer account documentation impacted the U.S. legal channel, causing a delay in our ability to file lawsuits. Then in early 2016, we began to feel the impact of our consent order, at the same time there were additional enforcement actions with other companies that impacted our industry. During this time, we also pursued U.S. productivity too far and led our collector staff attrit to levels where, in hindsight, we know the absolute profitability of our U.S. core portfolios had suffered. In Europe, we faced a different set of challenges, with a competitive environment and the establishment of operations in Italy.

I could not be more proud of the work our team has accomplished in response to all of these challenges. We identified the issues, began to move forward with solutions, adjusted operations and even reset our portfolios for the new environment. Our legal channel operation were back up and running in a few short quarters, and we quickly worked through the backlog of accounts. Our dispute team was completely build out, worked overtime, refining the process to be effective and efficient and is now adjusted to this new normal.

Our HR team worked nights and weekends to almost double the collector workforce in just a little over a year. Our current collectors worked unbelievably hard to keep in touch with our customers and servicing existing payment plans. Our European team established operations in 2 new markets, again, developing an SME business, and is being disciplined in their buying and is ensuring that operations in all countries are running smoothly. We are encouraged by the industry trends we are seeing. Supply in the U.S. is increasing. U.S. cash collections are showing positive trends in both core and insolvency. We are looking to hire even more collectors and add a new site to handle what we believe will be even more buying volume.

Also keep in mind that while all of this was going on, we settled 2 large lawsuits on reasonable terms, while also managing to increase both our European and North American credit facilities in order to have plenty of capital available to -- for the supply that we see on the horizon. As we consider the long term, I want to assure everyone that we will remain dedicated to the same principles we have long held: compliance, people culture and collaboration, data and analytics, execution and a long-term focus on success. It is crucial that everyone at PRA focuses on and values these areas, as they have been and will continue to be the keys to our success.

We are one company and one team worldwide. Operator, we're now ready for questions.

(Operator Instructions) Your first question comes from the line of Bob Napoli from William Blair.

Just a question, first, I guess, on Page 11, the $187 million base that you show. Now -- I mean you're suggesting that, that $187 million you then would add, like you did in the -- Page 4, the revenue from buying in the quarter and collections from fully amortized pools, less allowances. Then you would have yield raises on top of that. Is that the correct...

Yes, that's right. So -- I mean, the way the model works -- the way the accounting works is yield times NFR balance. And so everything that's going to happen this quarter is known when we filed this quarter's Q, except for those 4 things. So revenue that we would recognize on portfolios that we buy during the quarter, which we won't know until we close the quarter, as well as allowances or reversals that would happen during the quarter. Again, we don't know that until we close the quarter, and same for yield raises.

Okay. That's what I thought. But just on -- and then just the follow-up is on, I guess, on yields. Your thoughts on the yields on the paper that you are currently buying, and then I don't -- so you're suggesting from that earlier page that you didn't increase yields in the second quarter. So the yields on the paper you're putting on and then any yield -- the thoughts on yield raises or diminishment?

Yes. So again, I think, given our discussion around collector capacity and record volume of accounts being purchased, we're displaying some caution in both how we booked the deals that we're buying in the quarter as well as making sure that we don't raise yields too early. And so we're exhibiting a level of caution there until we get our hiring program completed.

You're next question comes from the line of David Scharf with the JMP Securities.

First, I guess, I wanted to just make sure I was interpreting the commentary on both the parts of the U.S. and European market. First on Europe. I was trying to reconcile -- I think the initial comments were you're starting to see more flow deals or longer-term flow deals, suggesting that maybe pricing might be stabilizing or at a peak. But I think you were also cautious that the near-term buying outlook is going to be lower than originally forecast. And I wasn't -- they seem to be dueling observations, and I just wanted to ask how we should think about that market for the rest of the year relative to where it's been lately? Is your sense that it's still very, very competitive and causing some near-term restraint on your part? Or are you feeling more confident that competition is peaking because of some of these longer-term flow deals you're seeing?

Yes. So in Europe, I mean, our opinion is that it is very high pricing right now. And the example I gave was trying to describe that from our perspective we believe that banks are entering longer and longer [probably] flow agreements because I think they, again, they believe that pricing is high. So that's our read on that. So generally speaking, pricing is high to irrational in a lot of areas in Europe. Now I'll tell you that we're watching this. We are watching all the results of all the public players and as well as anybody we can get our hands on the data and looking to see if there's any financial cracks. And just be aware that I'd love to help Europe out in their bad debt situation. Now, I also want to say though that this is not uniform. It's not uniform across Europe. This intense competition is everywhere, but I would say that in general if you look at the U.K. market, for example, with a little more mature regulatory environment, it tends to be a little more, again, mature, and I would not couch it as irrational.

Got it. That's helpful. And then switching to the U.S. Obviously, the Chapter 13 purchasing stood out, and that was always one of your differentiated sort of core competencies. Were you trying to suggest that there were 2 unique, very large transactions being worked on in the first half of the year, and we shouldn't necessarily draw conclusions about first half purchasing levels going forward? Or are you just seeing finally a big uptick in not only in supply, but given those OCC issues that overhang that market are you see some loosening up there of supply?

So for right now, the signaling or the very direct communication from our part is that 2 quarters does not a trend make. These were indeed 2 spot transactions from a seller. This seller had experience before selling, and they made an evaluation, and we brought a sharp pencil and a good relationship to the deal. So I would say that we're capable -- our insolvency or BK, as I call it in the States, that operation is ready and willing to purchase more bankrupt 13 paper. They're highly scalable and very motivated. But I'm not necessarily counting on it for the rest of the second half of '17.

You're next question comes from the line of Mark Hughes from SunTrust.

Did you provide specific numbers on allowances for the quarter?

That will be in the -- it'll be in the Q tomorrow when it's filed. But the total allowances for the quarter was a little over $3 million globally.

And then Kevin, you had talked about part of the reason you're adding capacity is because you see supply on the horizon. Is that essentially the rise in credit card charge-offs? Is there something you're hearing from sellers that gives you confidence they'll be selling more of the charge-offs or -- just sort of curious what you to look at when you describe plenty on the horizon?

Yes, thanks for the question, Mark. The -- from our perspective, both to service our existing portfolio and with the forecasted buying, certainly for the rest of '17 and on into '18, we think we need to make this move. Even more pointedly, if you look at the collector headcount, we're now over 2,200. And I don't know -- I didn't mention it on this call but I've mentioned it before that we have seating capacity for 2,200. So we're probably at about 105% of capacity. And just try coming around here some afternoon when people are coming in and out of parking lots, it's hard to find parking space. So we are -- we've been dedicated to a concept forever since our inception that we think collectors should have their own real estate. They should have their own desk. Real-estate prices where we locate call centers are not prohibitive. And so we don't like people to, as we call it, hot seating. And so based on the volume of existing accounts, based on the volumes we're seeing for the rest of '17 and '18, we think that opening 2 new call centers is the right decision.

Your next question comes from the line of Robert Dodd from Raymond James.

Just looking at the revenue yield -- and I appreciate the charts and the guidance. Obviously, you made a comment that some of the stuff that's rolling off is lower yield. Given you know how those portfolios and the vintages age much better than we do, where -- timing-wise, not the level, but just kind of timing, where would you expect those -- the gross revenue yield to bottom out? In terms of like next quarter, 2 quarters, 5 quarters given you know how those vintages age? I'm not asking for a level of the yield, but just kind of based on the mix of vintages today what's the timing of where that hits a bottom?

Yes, I guess, one thing I would say is I'm not sure if you intentionally misspoke, but the roll-off the portfolio is higher-yielding portfolios--

Right. Yes, yes, I did misspeak about that...

Not lower yielding. There's a lot of moving parts in that mix between the bankruptcy portfolios, between some of the acquisition portfolios in Europe and also the peak core buying. It's really hard to say exactly when that inflection point will happen.

You're next question comes from the line of Colin Ducharme from Sterling Capital.

Looks like he might have got disconnected.

I apologize. I had the mute button. Just curious, first question, if you could give just -- around any color whether on purchase multiples or pricing trends that you're seeing year-to-date, especially given the increased purchases you all foresee in the next quarter or 2?

I can give you some color on what we've seen over the past, call it, 18 months, and we can kind of go from there. So we've been talking about increasing or improving pricing. It's been in kind of a steady melt up over time. And again, this is nothing unusual or really nothing unexpected given this particular environment we're in. So we've just been on kind of a steady march along the way, trying to stay in constant dialogue with the sellers, by the way, and let them know what's going on. You're going to -- that's driven by a couple things. One is this, again, the volume issues and also the quality of the accounts. And as Pete mentioned, when you get the 10-Q though, you'll probably find that from a deal multiple standpoint Pete's booked them with a little bit of a haircut, really in line with what he's talked about earlier in the call.

Got you, okay. And then just a quick follow-up on Europe. You talked about things getting competitive. I'm curious just on color -- with some of the markets that have been smaller addressable market to date but have potential longer term, curious if you've seen any movement from the sellers in terms of their propensity to sell on the one hand? And on the other hand, from the other side of the market, in the hotter markets, where you've see some consolidation, I'm curious on the effect, if any, that you've seen on deal pricing as some of these markets begin to consolidate players?

So I think your question -- I was trying to tease it apart and take a couple notes here while you're talking. The question is -- I guess what I'll do is say is it a more pan-European pricing environment? Or are there still pockets from smaller sellers in Europe? I don't have that data. Tiku actually may -- Tiku is actually on the call as well, our CEO from Europe. And he might have some commentary on that. Tiku?

So yes, I would say that there are some pockets where the pricing is different and the returns are different because there are opportunities where sellers want to play to a particular proposition that we offer. And that's an advantage of being in the breath of countries and dealing with the breadth of vendors that we've dealt with over the past 15 to 20 years in Europe. So we try and play to that wherever we can. So we certainly do see some of those opportunities. But generally, the market is much more competitive right across the marketplace.

Your next question comes from the line of Bob Napoli from William Blair.

Just I guess a question on the amount of buying that you've done and the capacity, the lack of capacity and kind of the timing of running your capacity that tight where the increases -- I mean you're not going to be able, I think you said until October, increase. So the paper that you just bought is -- and I know you collect over a longer term. But just what was the thought process in not having -- doing the buying that you did without the capacity?

Bob, that's a great question. So I will tell you -- I'm getting some feedback from somewhere. So from our perspective, our capacity issues are, again, not only accounts we have in-house today but also projected buying. So let's make sure that that's out there. What we're doing is we are trying to change some of our scoring a bit and our dialing. And if you can imagine, one of the challenges that it presents to us, and I think you've hit it on head, is that as you buy a new portfolio and you've got a certain amount of capacity, from a score perspective a lot of times what happens is the newer accounts might score better, and they kick out the older ones. So we are making a very, very conscious effort to make sure we're getting fair representation of dialing throughout the portfolio. At the same time then, Pete is taking a bit of a haircut in a bit of a change in curve shape for the accounting models to make sure that we are dealing with those from an accounting perspective. And Chris -- Chris Grave's acquisitions group, they see this. Yet I remind -- make sure you know when we price something we are looking at the current capacities. So I don't know if that the answers to your question come or not? We might have -- Chris, you want to add to that? Chris is here as well.

Bob, one thing to think about too is the absolute size of our portfolio is so significant now that we get pretty good attrition just by collecting it in totality. So that opens up available capacity. So if you think about how much we need to buy just to replace that attrition based on the size of the portfolio, we get some nice benefit for handling our future capacity. And realize that we're not, obviously, collecting, to Kevin's point, the entirety of our new buys. Our data set is so large and our models allow us to get pretty surgical to pick out the top tiles of accounts to collect. And over time, those models get better and better and more and more surgical too.

Okay. And just to be clear, the revenue yield on the older bankruptcy is higher, like when you go back several years, than what you're putting on today from those very high yields. But then, the America Core that you're buying today should be comfortably above the average yields on the portfolio today. Is that correct or not?

So let me take first crack at this. You're absolutely spot on with the bankruptcy -- U.S. insolvency portfolio. Some of those yields were truly exceptional. I think most of us on the call probably remember the history of the cram-down legislation threat that drove some of that. The core though I would say core -- I don't have the analysis in front of me. But since I was deep into level yield accounting for so many years, I would say that we are indeed running off higher-yielding portfolios. And so, Pete, might want to add to that?

Yes, and I think there's a nuance there to keep in mind is that there's a difference between underwritten yield and the revenue-recognition yield. So naturally as a portfolio ages, the yield goes up because of expansion over time. And so the yields that we're recognizing revenue on, some of those older portfolios are significantly higher than the base underwritten yield would have been. And the offset to that being newer portfolio that we're putting on and, again using my words not Kevin's, booking in a more cautious nature in the near term, so booking at a lower yield -- lower level.

You're next question comes from the line of a David Scharf from JMP Securities.

Just couple of follow-ups. One real quickly. On the tax rate, Pete, going forward, was this a catch-up item, the 1.9? Or based on the geographic mix of income going forward, should we be looking at the consolidated rate moving above that 36%, 37% prior guidance?

Yes. So if you look at our year-to-date tax rate, it's 39.9%. And so that onetime impact that we highlighted for the quarter was us readjusting and taking a half-year adjustment in the second quarter. And so that's kind of how you get to that item that we called out. What I would say in terms of overall tax rate, our base U.S. tax rate wants to be, call it, 42%, 43% when you factor in state tax on top of the federal. So as we shift to having a higher proportion of our earnings coming from the U.S. business, that's going to tend to trend in that direction.

Got it. Got it. And then maybe a little bit of follow up on Bob's questions regarding the call center ramp. Should we be thinking about any sort of step function in fixed expenses sometime in Q3 or overall second half as you build out these centers? Or is it just pretty modest, a lot of it's capitalized? And soon as people are in their seats, it's mostly variable expense?

Yes, David. That's correct. So I think just roughly speaking it cost us, what, $2.5 million-ish, somewhere in there, to open a call center. And most of that is capitalized from TIs, account improvements through furniture and computers and so on. So that's that. The only -- Pete talked about some more sluggishness in terms of cash efficiency ratio simply because we're ramping up all those people. And that's kind of on your point. On average, we're diluting our workforce since we're adding so many folks, and we need those guys to mature. The rent itself isn't a big deal. We're not paying New York City rental rates. We tend to be in areas where rent is very affordable. And again, one of our key premises is we want to give collectors a nice place to work at, in their own space and that -- we're able to do that.

You're next question comes from the line of Bob Napoli from William Blair.

Just a quick last one. Any -- what's going on in Brazil? Are you buying much paper in Brazil? Can you give me -- give us some idea of what's going -- I mean, the buying that you're doing and the outlook for the Brazilian market?

So, Bob, yes, just in general, from a color perspective, we're doing well in Brazil. We like the market. We love our partners down there. And -- but the buying is not huge, and it's lumpy. So right now, I don't think we see any more flow of capital down there. And because we injected the capital initially, I think they still got decent amounts to invest. And when they send me a deal memo, I like to approve it generally.

You're next question comes the line of the Colin Ducharme from Sterling Capital.

Just a quick follow-up for either Kevin or Pete. Just curious, because as you say, Kevin, you guys are keeping a sharp pencil and from what I understand about the culture very metric focused. But I'm looking at a press release which gives us less information then we have kind of long had, specifically regarding kind of returns on equity, returns on assets. You used to have return on equity displayed at the top of the release, and the table doesn't appear in the most recent release. So just curious what signal if any to read from that and the thinking going behind that if there's a focus away from those types of metrics henceforth?

That's a good question. No, that was my doing. What I've tried to do is the data -- that very busy table in the back of the press release is all data from our 10-Q's and 10-K's. So we -- instead, we tried to give you guys a little more color in the press release and commentary, more in line with what we're talking about here this evening. So we're trying to morph the press release more into something that's interesting maybe to read and a little less forensic. Don't take anything away from that. Our focus on data and analytics and underwriting and all that remains unchanged. But we're just trying to make the press release a little more, I think from my perspective, a little more helpful. And then when the 10-Q or 10-K is filed, you'll have all that data.

You're next question comes of the line of David Scharf from JMP Securities.

Just one more. Staying on the call center productivity theme. Obviously, compensation's your biggest expense line item. How is the -- just curious how call center turnover is currently maybe in comparison to 12 and 24 months ago. We're at, obviously, 4.3% unemployment, arguably full employment. Any increases there, stresses? Or has it been pretty stable throughout the cycle?

David, I actually have that data for you. I was counting on someone to ask that one. So the answer to your question is it is definitely up. When you're hiring this many people, you expect turnover to be up. What I think is interesting -- and I'm going to give you the numbers in a second. What I think is interesting is the current turnover rate, again, this is simple math from our Human Resources Department, is about 68% turnover rate. And that's for call center collectors, and that's from day 1 of hiring. That is not an unusual rate for us, by the way, historically. So you've been around long enough, you remember those rates, 60%, 70% nothing unusual. What I think is fascinating about it is that we've talked for the past, I don't know, year that we let our collector force attrit. And what you end up doing is keeping your very best collectors, generally. And so our turnover rate actually went -- I've got it here from 2014 on. So in 2014, it was about 52%. In '15, it went to 38%. And then it went to 42% in 2016, and now it's back up to 68%. And at the same time, collector productivity, using the round numbers, went in '14 -- in 2014, it was about $84 an hour paid. Then it spiked to $115 to $143, and now it's about $94. These are round numbers. So it's an interesting study. But I would say what I'm pleased about right now is that our turnover rate is nothing unusual compared to our historical numbers.

Got it. And then lastly, as we think about margins going forward, given that legal collections are a more expensive option if you don't collect out of the call center, is the greater placement -- is the channel mix shifting more weighted towards call center? Is this just a function of where we are the cycle, with a healthy consumer you don't have to sue as much? Or is there more of a concerted effort to not rely as much on the legal channel just because of the new documentation requirements and the margins and so forth?

Wow. So let's dissect that one, and I think it's a good question. So first of all, the way you ended it was the new documentation requirements. Again, the documentation requirements are coming from the OCC and the CFPB, not necessarily the courts. So we have more documents today than we ever have. So theoretically, you would argue that actually legal is easier for us because I've got these documents. I don't have to order them. They're already in-house. So -- and I'll just -- I'll also say this. I think about that constantly. And I think about all these docs are in my hand. I've got a consent order. I've got other consent orders. I've got requirements, and I always think about any unintended consequences that might be buried inside those. So let's just say that the book's always open, looking at these returns. But back to your question. The reason that we see these accounts moving more into the call center as opposed to legal is really balance driven. It's legal ROI driven. If you think about your investment in an account -- so remember, we want to file lawsuits against people who are won't-pays. So to the extent that you have to invest -- double down, so to speak, and invest in court filings and everything else to process a lawsuit, if that balance is smaller it gives you less room for any kind of error. So just -- all things being equal, lower balance tends to drive less legal and tries to push it more towards the call center. Again, I always have to backpedal though just to let you know that we're always looking at that legal score and that legal ROI. Because again, we're trying to file lawsuits against people who are won't-pays, not can't-pays.

I am showing no further questions at this time. I would now like to turn the conference back to Mr. Kevin Stevenson.

Well, thank you very much. Thank you, everyone, for attending the call. We look forward to speaking with you next quarter.

Ladies and gentlemen, this concludes today's conference. Thank you for your participation, and have a wonderful day. You may all disconnect.