Navient Corp (NAVI) 2017 Q3 法說會逐字稿

Good morning. My name is Andrea, and I will be your operator today. At this time, I would like to welcome everyone to the Navient Third Quarter 2017 Earnings Call. (Operator Instructions)

Mr. Joe Fisher, you may begin.

Thank you, Andrea. Good morning, and welcome to Navient's 2017 Third Quarter Earnings Call. With me today are Jack Remondi, our CEO; and Chris Lown, our CFO. After their prepared remarks, we will open up the call for questions.

Before we begin, keep in mind our discussion will contain predictions, expectations and forward-looking statements. Actual results in the future may be materially different from those discussed here. This could be due to a variety of factors. Listeners should refer to the discussion of those factors on the company's Form 10-K and other filings with the SEC.

During this conference call, we'll refer to non-GAAP measures we call our core earnings. A description of core earnings, a full reconciliation to GAAP measures and our GAAP results can be found in the third quarter 2017 supplemental earnings disclosure. This is posted on the Investors page at navient.com.

Thank you. And now I'll turn the call over to Jack.

Thanks, Joe. Good morning, everyone, and thank you for joining us today and for your interest in Navient. My comments this morning will cover my perspective on this quarter's results, the outlook for future growth and how we plan to increase the enterprise value of Navient.

Our adjusted core earnings of $0.56 was rewarding and better than forecasted. The results reflect solid performances across the board as our net interest margin, credit performance, fee revenue and operating costs combined to contribute to the quarter's strong results. This quarter's results included some onetime items. Excluding these, core earnings were a strong $0.52 per share.

Highlights from the first 9 months of 2017 include the acquisition of $8.6 billion in student loans. We completed the acquisition of Duncan Solutions. We executed an agreement to acquire Earnest. We increased our noneducation-related fee revenue by 29% year-over-year. We reduced our 2018 unsecured debt maturities to less than $1.4 billion, and we distributed over $570 million to shareholders. In addition, student loan performance has been a bright spot for both private and federal student loans. Delinquency and default rates continue to decline for both types of loans as we continue to deliver superior performance amid an improving economy.

Earlier this month, the Department of Education released the results of the 2014 3-year cohort default rate. While the national rate increased, the rate for Navient-serviced loans declined. Navient-serviced borrowers were 37% less likely to default than borrowers serviced by other organizations. We also continue to be a leader in enrolling federal student loan borrowers in alternative payment programs like income-based repayment. In fact, 53% of the loan balances serviced by Navient for the Department of Education are enrolled in income-driven repayment programs, more than any other servicer.

This is all good news. Yet, despite our consistent industry-leading performance in assisting federal student loan borrowers, the baseless, politically-driven accusations have continued. I'd like to take a few moments to discuss one such example, the recent lawsuit brought by the Pennsylvania Attorney General. The Attorney General asserts that he brought charges after an extensive investigation, but no one from his office has ever stepped foot in one of our centers or requested information on Pennsylvania borrower accounts. Instead, they simply copied the lawsuit filed by another state.

He claims to have received thousands of complaints from Pennsylvania residents when, in fact, 53 Navient-serviced borrowers requested assistance from his office in the last year, and 49 of them reflected disagreement with federal policy or a term of the loan, not a concern regarding our servicing. He claims we steered borrowers into schemes where we "jacked up their interest rates." Yet, federal student loan interest rates are set by Congress and are fixed and are never changed by the servicer. And once again, there is no claim that we violated the terms of the Higher Education Act or our servicing contract. We've done a cursory review which showed that there is a stark chasm between reality and these unfounded claims, and of course, our industry-leading performance statistics stand as a factual contradiction to these accusations.

Perpetuating baseless allegations for political gain does nothing to assist borrowers towards successful repayment. In fact, it's more likely to harm borrowers by directing attention away from the true issues that are causing student loan defaults: a lack of adequate upfront education, an overly complex repayment system and the serious financial consequences of dropping out before graduating or not responding to servicer outreach. We should be focusing on addressing these issues.

The federal loan programs are complex, too complex. Ironically, the solution promoted by many would increase this complexity. Struggling borrowers have difficulty navigating today's complex set of rules and need help to understand their options so they can make the best choice given their circumstances. Year after year, we demonstrate that when we speak with a struggling federal loan borrower, over 9s and 10, we help them select a solution that keeps them out of default. This is what we do, and it's why our customers default at a 37% lower rate. In the meantime, our repeated request that regulators encourage struggling borrowers to contact their servicer goes unanswered.

It may be -- it may appear to be politically beneficial to make claims of widespread abuse, but the facts will rise to the surface and I'm confident that the facts will be heard and the extraordinary results delivered by Team Navient will defeat this politically-driven campaign.

It is, however, disappointing that we have to spend time responding to these. But in spite of this noise, our team continues to provide outstanding support to our customers and has remained focused on creating value in each area of the business. What we are doing to create value is the main focus of my comments this morning.

In our legacy student loan business, we've been creating value from student loan portfolio purchases and maximizing cash flows. Since the separation in 2014, we have acquired nearly $30 billion in student loans, adding almost $6 billion to projected cash flows. This is real and very tangible value for shareholders. We've also generated growth and delivered value from our fee-based businesses. We have acquired several companies, including Duncan Solutions this quarter, to expand our footprint and help drive organic growth. Our services leverage our operating skills and systems to business partners beyond education loans. We are building a business here that will generate double-digit organic revenue growth with very attractive risk-adjusted margins.

Finally, I believe we can generate meaningful value by originating new loans in addition to purchasing legacy student loans. Today's student loan programs use a one-size-fits-all pricing or charge rates that reflect the risk of a student not graduating. With over 40 years of experience, we know that borrowers who graduate gain the benefit of their investment in education with higher levels of employment, higher incomes and stronger financial health. Our new loan products are focused on helping young adults refinance their student loans at the lower rates they have earned. For these customers, payment performance is extremely strong and defaults are very low.

To help accelerate our pace of growth in this area, we agreed to acquire Earnest, a major player with a distinctly different approach to the market. They have developed solutions that deliver exceptional customer experience, dynamic pricing and the ability to use data to create value, all at a lower cost. The result, they have earned the highest Net Promoter Scores in the industry. I believe our combination of our respective skills and resources will create more opportunity and value, significantly more value than either of us would alone. I'm excited about the opportunity to build a great business with our newest colleagues and to create meaningful value for our customers and our investors.

We will add over $1 billion in new refi loans to our balance sheet this year, and I expect we will add more than $1.5 billion in new loans in 2018. I believe we'll be a dominant player in the space, generating high-quality loans, producing mid-teens ROEs. We've now completed or will complete several transactions in 2017 that have and will create significant value for our investors. Specifically, we will add over $10 billion in student loans to our portfolio. We added Duncan Solutions to our fee-based businesses, and with Earnest, we'll be a dominant originator of refi loans.

These additions require capital. To maintain a strong balance sheet and financial flexibility, we suspended our share repurchase program through 2018. I believe this represents an investment though that will reward shareholders. First, we're reinvesting this capital for growth. I believe we'll generate very attractive low-risk returns with projected ROEs from these investments in the mid- to high teens. And second, our investments will position the company to generate earnings growth. I believe it will be earnings growth and the clear visibility of earnings growth that will attract investors and allow the company to be fairly valued.

As shareholders, you know that this has been a significant challenge to date and even more so today. My job as CEO is to create value for our investors by delivering great products and services to our customers and clients. I believe we've done this, but as we all know, it's not reflected in today's stock price. My team and I are confident that our strategy, our products, our services and our people will create value in the returns you and I believe are possible. I'm confident of our ability to deliver on this potential and create value for our investors.

I'll now turn the call over to Chris for a deeper discussion of this quarter's results and guidance, and I look forward to your questions later in the call.

Thank you, Jack, and thank you to everyone on today's call for your interest in Navient. Throughout this call, I will be referencing the earnings call presentation, which can be found on the company's website in the Investors section.

Starting on Slide 3. We reported adjusted core EPS of $0.56 in the third quarter compared to $0.51 from the prior year. During the quarter, we recognized $47 million or $0.11 a share of deferred revenue related to loans for which Navient will no longer provide services due to the previously announced termination of the related services contract. This deferred revenue is recognized to reflect a shortened period over which it is expected to be earned.

As we have discussed in previous quarters, we have seen an increase in prepayment speeds as the economy has improved and our portfolio continues to season. We therefore adjusted our prepayment speed assumptions, otherwise known as our CPR assumptions, in the third quarter to reflect this activity. As a result of this adjustment, there was a $28 million net reduction to net interest income.

We periodically assess and, if necessary, adjust prepayment assumptions used to amortize premium and discount on student loans. Our prepayment speed assumptions used in our cash flow projections have historically ranged between 3% and 6% since 2010. Year-to-date, our core earnings per share excluding regulatory costs is $1.36. We expect fourth quarter earnings per share to be at least $0.40. This includes the impact of the acquisition of Earnest and the temporary suspension of toll road services in Puerto Rico.

Let's now move into our segment reporting, beginning with FFELP on Slide 4. FFELP core earnings were $46 million for the third quarter of 2017 compared with $69 million in the third quarter of 2016. The net interest margin for the third quarter was 71 basis points compared to 87 basis points a year ago. Excluding the aforementioned CPR adjustment, net interest margin would have been 86 basis points. This quarter, we continue to enter into derivative transactions to further mitigate the potential dislocation between 1- and 3-month LIBOR rates. We anticipate FFELP NIM for the fourth quarter to be in the high 70s.

In the third quarter, we acquired $523 million of FFELP loans, bringing our year-to-date FFELP acquisition total to $5.2 billion. Credit quality on the FFELP portfolio continues to improve with the increase in forbearance levels, primarily a result of the disaster forbearance that was granted in the quarter. We expect to see continued improvement as this portfolio continues to season.

Let's now turn to Slide 5 and our private education loans segment. Core earnings in this segment remained flat year-over-year. In the third quarter, the net interest margin was 357 basis points or 348 basis points excluding the CPR adjustment. The 29 basis point increase in the net interest margin from the second quarter was primarily due to a full quarter's impact of the $3 billion education loan portfolio that was acquired at the end of the second quarter. We anticipate the private education loan NIM for the fourth quarter to be in the mid- to high 330s. This updated guidance includes the anticipated acquisition of Earnest assets and the market's anticipation of a rate increase. Private education loan losses and delinquencies continue to decline as the portfolio amortizes. The year-over-year improvement was expected, and charge-offs are at their lowest levels since 2008.

With our FFELP portfolio, we saw an increase in forbearance as a result of clients that were granted disaster forbearance in the quarter. Our reserve policy provides for a minimum of 2 years of expected loan losses so we do not expect to see any increase in provision related to these impacted loans. As a reminder, we expect charge-offs and provisions related to the acquired $3 billion education loan portfolio to increase in the fourth quarter. This portfolio was purchased at a discount last quarter, and we will report provisions for losses as charge-offs occur.

We will now turn to Slide 6 to review our business services segment. In this segment, core earnings were $105 million in the quarter compared with $81 million in the third quarter of 2016. As Jack previously mentioned, we acquired Duncan Solutions in the quarter. This acquisition, along with organic growth, led to a 29% increase in noneducation fee revenue to $56 million. Duncan Solutions will be highly complementary to our Gila platform and create attractive growth opportunities. During the quarter, we also announced that we will provide portfolio management and post-default collection services for the combined Great Lakes and USA Funds FFELP portfolio.

Let's now turn to Slide 7, which highlights our financing activity in the quarter. In the third quarter, we purchased $648 million of education loans. We have available capacity in FFELP ABCP facilities of $3 billion and available capacity in our private education loan ABCP facilities of over $600 million, including our most recent private loan securitization. We plan to fund our education refinanced loans with ABCP facilities before we securitize them to term.

We issued 2 FFELP ABS transactions totaling $2 billion. Year-to-date, we have issued $5 billion of FFELP ABS, with the most recent deal financed at an all-in cost of funds that represented a 30% improvement over a year ago. On October 12, we priced our first private education loan securitization since February 2016. This $662 million transaction is our first deal to include a combination of recently originated education refinanced loans and legacy education loans. This transaction price had a spread of 101 basis points with a weighted average life of 3.5 years.

Moving to Slide 8. I want to quickly review our unsecured debt maturity schedule. In the quarter, we reduced our unsecured debt maturities by $548 million. In the past year, we have reduced our 2018 maturities by 33% to $1.4 billion.

Since our separation in May 2014, Navient has returned $3.3 billion of capital in dividends and buybacks to shareholders. In addition, we have acquired nearly $30 billion of student loans, purchased highly complementary fee-based businesses for $435 million and agreed to acquire Earnest to accelerate Navient's ability to capture a significant share in the large and growing education financing market. We believe all these transactions -- these actions have increased the intrinsic value of the company.

As a result of these positive actions since our separation, we also saw our TNA ratio decline to 1.22x in the third quarter, which is in the bottom quartile of our stated range of 1.2 to 1.3x. We have therefore temporarily suspended our share buyback program to fund increased originations expected from Earnest as well as to build capital to increase our TNA ratio. As we execute on the aforementioned initiatives over the next year, we expect to reinitiate our buyback program in 2019 and we will maintain our dividend policy.

Finally, let's turn to GAAP results on Slide 9. We recorded third quarter GAAP net income of $176 million or $0.64 per share compared with net income of $230 million or $0.73 per share in the third quarter of 2016. The primary differences between core earnings and GAAP results are the marks reflected to our derivative positions.

I will now open the call for questions.

(Operator Instructions) And your first question comes from the line of Sanjay Sakhrani from KBW.

I guess my first question is on the FFELP NIM. I guess the guidance is that NIM will come down into the high 70s from the high 80s. And I was just hoping you could just sort of, Chris, maybe walk us through what will drive that. I understand you guys entered into certain hedges, and you also are expecting rates to go higher. But aren't the hedges supposed to mitigate some of that impact? Or is that the cost of the hedge?

No, it's a very good question. And I want to provide some color on that guidance, so thanks for asking it. So what we inevitably do in providing the guidance is provide a base case for NIM guidance that we believe -- that we feel confident in coming through. There is some variability in the financing of our FFELP portfolio, and inevitably, at times, that works to our advantage. And we've seen some of that advantage over the last few months, but the guidance we provide is assuming that, that variability ends up on the lower end. The hedges of the portfolio actually are working just as we expected. They have reduced the variability in the 1s, 3s balance. In fact, today, roughly 94% of our 1s, 3s position is hedged at an all-in cost of about 10 basis points versus the cash cost today of about 11 basis points. So the hedging program is doing exactly what we had hoped it to do in reducing that volatility. So it really has to do with some variability on the financing side. And inevitably, when we end up through our guidance, it usually is a result of that variability working towards our advantage. Does that makes sense?

Sort of. And so I guess, when we look ahead and assuming -- I don't know what the forward curve assumes, but like when we think about the FFELP NIM into next year, should we expect that to navigate back into the level that we saw in the third quarter? Or does it stabilize, in theory, if there's no rate change obviously, in the high 70s?

So there are obviously a number of components around the FFELP NIM. But in the -- so the 1, 3 spread which we've hedged, which reduced the volatility, is something to make sure to keep track of. Obviously, floor income is another important component, how that gets brought into our NIM year-over-year. And as you know, that's an amortizing asset so there will be less of it every year, and that's an important component. But inevitably, we do strive for stability around that FFELP NIM, and we will provide guidance at the end of the fourth quarter into what we expect into 2018.

Okay, great. I guess my second question is on Earnest and the mid-teens -- you guys talked about the mid-teens ROEs. Maybe Jack or Chris, you could talk about the ROAs expected in that business and sort of how you get to those mid-teens ROEs.

Sure. So we look at these loans as having the potential to generate kind of net interest margins in the low 200s. And when you look at the high average balances associated with the accounts, the servicing costs are a lower percentage of the overall mix. And as I said in my comments, when you're lending to borrowers in this segment, you're lending to borrowers who have completed their education and have demonstrated income through their existing job and a very stable, consistent payment pattern early in their repayment years. So default rates in this space are extremely low. You can look at some of the public securitization transactions that have been done in the space, and credit losses cumulatively are in the low double-digit basis points.

So when we think about the ROAs, I mean, should they work out somewhere around 1% or a little bit lower than that then?

It should be a little bit higher than that.

Okay, a little bit higher than that. Okay. And then final question on the Puerto Rico impact that you guys talked about. What is explicitly that impact as we look in the fourth quarter? And how long will that persist?

So with the hurricane that occurred there, the Commonwealth suspended toll collections. We run -- we manage the toll roads for the Commonwealth there. That's a temporary suspension. They're starting to -- they're now getting back up online and have begun collecting tolls again. Not all of the different booths are open at this stage in the game, but we're expecting that, by year-end, we'll be back to full operations across the island. So the -- I mean, the impact is...

And the revenue in the fourth quarter.

Yes. Total revenue -- there's a fair amount of variable operating expense that goes against that revenue, but it is -- which is mostly contractor expenses. So it was expected to contribute maybe $2 million of net revenue in the quarter -- fourth quarter.

I think importantly to that is the fact that they are operational again. I think -- we think about it's roughly 50% to 60% back to operation and it's October. So there is a strong expectation we'll be fully back online, but clearly, there was disruption from the hurricane.

Your next question comes from the line of Michael Tarkan with Compass Point.

Just back on Earnest and the ROE expectations there. Can you just tell us how much capital you plan to hold against Earnest loans and then how much -- as we think about 2019, how much capital you plan on putting towards in-school loans?

So the capital targets that we have established for refi loans is really a function of what -- of the credit loss models that we've built and the stresses that we put against that. So we are targeting 5% capital against that portfolio given the exceptional performance of the assets. I think in-school loans -- lending in the in-school arena does require higher capital because you are lending with the graduation risk or graduation success unanswered at that point in time. But in the past, we have typically said somewhere around 8% for those assets when we were a different company. We will assess that if and when we get into that space in 2019.

Okay. And then just following up on that one. So if it's 5% against Earnest and you plan to do over $1.5 billion, I guess what I'm trying to get at is the suspension of the entire buyback for 2018. 5% against $1.5 billion is $75 million. I'm just kind of curious how you came to that decision, if you can give some color there.

Okay, yes. So I think that's a great question, and I don't think you should be looking at the share -- the suspension of the share repurchase program as limited to the Earnest transactions. We completed a number of transactions this year that consume capital. So we acquired $10 billion worth of student loans during the course of the year. Over $4 billion of those will be private student loans, for example. We acquired Duncan Solutions and Earnest, and we've set for ourselves a conservative tangible net asset ratio guidelines of between 1.2 and 1.3. We were getting towards the lower end of that, and our plan was to -- through the suspension is to rebuild that capital ratio pretty rapidly during the course of 2018 and give us the resources to fund the asset generations in this area. The company still generates excess capital even with our planned investments in growth, and we continue to maintain our dividend. And as Chris said, we'll -- we plan on reestablishing our share repurchase program in 2019. But it's really about a prudent allocation of capital and where we think we can generate good returns. And clearly, in our view, we believe these investments generate attractive returns for investors that are not kind of onetime but are building businesses that will repeat those returns year after year.

Understood. And then last one for me is -- so with $1.5 billion originations out of Earnest in 2018 or more than that and then in-school origination starting up again in 2019, at what point should we expect sort of earnings to sort of bottom out and begin to start growing again? I know you don't want to look out too far, but just kind of curious if 2019 is a year where -- even absent the buyback, where we can see earnings dollars bottom out and start to grow or if it's a little bit later than that.

So I think the one piece that you need to -- that we need to keep in mind here is economic value of the assets being generated versus GAAP or our core earnings items. Originating loans have different -- have accounting issues that cause the asset to -- cause new loans to be -- that originated in that year to have a GAAP loss. And that's primarily related to the fact that your earning asset is on the books for less than a full year, but you are booking or expensing 100% of the origination costs associated with that loan. And in our loan loss reserve model, 2 years of projected credit loss is booked day 1. So there's some dislocation as a result of that, but we are clearly -- we clearly look at ourselves in these businesses that we're building in both the fee-based business and the asset generation side of the equation to be generating earnings growth in the future. I think when we talk later in January of next year, we're providing a little bit more guidance as to when that would happen.

Your next question comes from the line of Arren Cyganovich with Citi.

I guess just touching on the capital subject again. Clearly, you'll be able to build capital through the next year. I'm just wondering what the decision was -- I get the understanding you need to finance some of these new investment opportunities, but at the same time, you've been running pretty close to the 1.2 for several quarters. Was there a push from your -- from the rating agencies, from your bondholders, et cetera? What was really driving that decision to have to bulk up your TNA ratio?

So it's a great question, and it's obviously something we spend a lot of time thinking through. Inevitably, we have to manage a number of constituents in our position as a finance company and the rating agencies, the bondholders and the equity holders. I think we mentioned in our prepared remarks, since the spin of the company, there's been significant amount of capital allocated to shareholders, whether it's been in the form of buybacks and dividends, whether it's been buying assets without any impact on our buyback program. And so we just thought, from a prudence perspective, that we've gotten to a point at the lower end of our guidance or our range that given our financing needs over the next few years, given the capital that we will need to consume and given a desire to be prudent over the next year, we felt it was the smart thing to do just to pull back for 1 year to suspend the buyback and then reinstate it in 2019. So it was really more of a holistic view of the company and making sure that we manage all of our constituents in a way that we felt would produce the best outcome. But there was no explicit...

And there were no explicit discussions from the groups you mentioned.

That's right.

Okay. And there were no liquidity needs driving that, right? You clearly have enough excess capital to pay down your -- and push out your unsecured debt?

We feel very comfortable about our financing position.

Okay. And then just moving on to the -- I guess, the longer-term refi loan projections that you provided. The life of loan loss expectation of 1.5%, again, it makes sense from a high-quality customer standpoint. The low numbers you referenced in the recent transactions haven't really gone through credit cycle yet. Do you have any kind of historical expectation or any kind of testing around that to think about how that would flow through if we went into another recession?

Sure. So firstly, I should have said "thank you" for picking up coverage of the company. We appreciate that. So we have 40 years of experience here, and that 40 years of experience of how different types of borrowers perform in different types of credit cycles is one of the real distinct and unique assets of this company. We know from our experience in history through all credit cycles that borrowers who graduate particularly with 4-year or graduate degrees have substantially higher levels of employment, higher levels of income, and those persist even during periods of stress. So if you go back to the last recession, for example, where unemployment rates approached or hit double digit, hit 10% levels at their peak, the unemployment rate for all college graduates was half that. And when you're looking at what we're talking about lending to here, we're lending to even just a portion of that segment, where credit performance has been -- has really been exceptional across different credit cycles. That data, that experience in understanding what drivers or characteristics of each individual borrower produced those results, that is really the distinct advantage that we have that we think we bring to the refi space.

Your next question comes from the line of Lee Cooperman with Omega Advisors.

Yes. This is at a higher level, and I have to say I have to be somewhat critical. I think it was roughly 3 years ago, it's amazing how time flies, you opened up a conference call by saying -- by addressing the buyback, saying we agree with those analysts that have a liquidating or run-off value in the low 20s, okay, and that buying back stock was very accretive to the shareholders. Now I happen to be one of the shareholders that wasn't selling into the buyback, but you have bought back 167.4 million shares since 2014 at an average price of $15.17. The stock is now $12.20 or 20% lower than your average buyback. I would say roughly if you thought the value was in the low 20s before the buyback, the amount of stock you bought back at this price -- average price of $15.17 would probably add conservatively $3 or $4 to the value. So we're looking at a business that's worth, if you're -- unless you change your mind. I'm prepared to listen that you changed you mind that you were wrong because we all make mistakes in life, nobody's perfect, but it's probably a mid-20s value. You then entered into a transaction which really the market doesn't understand and doesn't like, dilutes the earnings and the stock goes down and we suspend the buyback. If we liked it, going back -- and I have the actual price you paid. And if we liked it at $15.17, we should like it a lot more at $12.20, and the debate that should be taking place at the board level is whether we should put the company into runoff and return the money to shareholders because there's nothing we could do in building and running the company to give us a value that we think exists if we put the company into runoff. So it's very disturbing to me that we buy back all the stock, and now that it's a lower level than we paid at any time the last 3 years, you're basically impotent and can't do anything. So it's frustrating. It means that there's been a poor job of capital management. So you want to address that?

Yes. And I take your comments in a constructive way, Lee. I know that's how you mean them too, so I appreciate it. Thank you. So I -- we do agree. I still believe that the value of this company -- the intrinsic value of this company is in the low to mid-20s per share, and we get there by looking at the discounted cash flows coming off of the legacy student loan businesses that we either own or run as a fee-based business and a combination of some minimal value to the ongoing fee-based businesses. So you can get to a very high number, a number much higher than today's stock price without providing some forward multiple on components of the growth businesses. Our job had been -- and this is how we -- I view our allocation of capital, is that we are supposed to be investing and using that capital that we are generating to invest in high-returns activities. We've done that principally to date by buying legacy portfolios. That is less of a franchise business and just adding to the existing cash flows. We've acquired nearly $30 billion. And as I said, that $30 billion worth of student loans, we expect to generate undiscounted cash flows net to share -- net after debt -- structured debt costs of nearly $6 billion. We have been building our fee-based businesses and now allocating capital to generating assets. I think our challenge really has been that, as we have been buying back shares, that -- I think you will recognize that value and what the intrinsic value of the company is. Many of our shareholders do. But attracting new shareholders and getting paid for that has been more of a challenge. And we think we can -- we have retained that intrinsic value of the company, and we can add to it by building these fee-based businesses and generating returns on equity that, I think, are pretty attractive in the financial services space. I mean getting mid-teens ROEs on very low-risk assets is much higher than what you see from most financial institutions these days with a whole lot less risk. So I think you used the word frustration. We are equally frustrated with the stock price, and we are working hard to invest in business activities and franchises that will hopefully get that share price or we believe we'll get that share price fully reflective of the value that we think is there.

Well, I guess the question I'm raising is how long does it take you to achieve that goal and whether there's a more expedient way of going -- look, your regulators crap over you constantly even though I think your business is a noble business. I personally sent 5 of my kids to college, and I paid their tuition, okay? And it's proven that an individual with a college education has earning capacity over their lifetime of well over $1 million more than a noncollege graduate. You have a noble mission, but somehow these politically-motivated regulators are constantly crapping all over us. So I say why don't we just take our bat and our ball, go home, go into a runoff mode, give the $25 back to the shareholders and let them decide how they want to reinvest their money. Because the question you should be asking yourself is if you really believe the value is in the mid-20s, what do you have to do to get there? And is it possible or practical to think that you can get there through operating the company or you should take a different approach? And I only raised it because, frankly, if you didn't make mistake in the buyback and we're 20% below the average price, and keep in mind some of these prices were paid 3 years ago, and you think the intrinsic value of the company is greater today than it was 3 years ago, we should consider taking a different path. But I don't want to monopolize the call. I just -- I'm supportive. I like you. You're a high-quality guy, but there's something wrong. That's all I can say.

All right. Thank you. Thank you, Lee.

Your next question comes from the line of Rick Shane with JPMorgan.

I just want to sort of get a sense of what the trajectory -- or how we should be thinking about the trajectory on the asset recovery revenue. This quarter, when we take out the accretion or the early recognition, it's back down to $110 million, which is essentially flat from the previous quarter. I'm curious, on the deferred revenue that was being recognized, what was the schedule there so we can understand how much was sort of taken out going forward? And then also, you talked about $55 million of annual revenues from Duncan. How much was recognized in the third quarter? And so should we just start to see a $14 million incremental run rate there going forward?

So a bunch of questions in there. Let me first address the deferred revenue-related issue. So we ran a business for guarantors, where we had agreed to provide services over the life of a federal student loan asset that was guaranteed as part of that process, and that -- the services were default prevention-related services. With the termination of that contract, we -- instead of recognizing the revenue over the life of the loan, we now recognize it over the remaining term of the contract, and so that's what caused the recognition-related issue. Many of the services that we are providing under our business processing solutions group are fee-based revenue services. So we're doing processing-related work, billing and payment processing activities and a little bit less asset recovery or collections services. Duncan falls more into that category of fee-based services. The deal closed on -- at the end of July, so there's roughly 2 months of revenue in there. And total revenue was $8 million for the -- recognized as $8 million for the quarter, the run rate.

Okay, great. What was the quarterly run rate on the terminated contract just so we can adjust that properly?

From a revenue perspective?

Yes.

Roughly about $4 million.

Per quarter?

Yes. I think it was probably a little bit higher than that, maybe about $7 million per quarter total.

Okay. So that contract had a couple of years remaining...

From all of the guarantors, I think ...

From all the guarantors.

Yes. I'm sorry?

So that contract had a couple years remaining on it when it was terminated?

Well, you -- it was more of a perpetual contract. So until the obligations, you no longer provide services, on outstanding guaranteed loans ended, you had to amortize it over the effective life of the loans itself. It's -- effectively, we're -- we got paid a fee, and we amortized that fee on an effective interest method.

Your next question comes from the line of Moshe Orenbuch from Crédit Suisse.

You guys had mentioned something about prepay speeds kind of changing. Could you talk a little bit about where they are now and what's embedded in the forecast for the cash flows?

Sure. So as we've been discussing over the last few quarters, our CPR speeds have been coming in higher than expected, and so you've heard, over the last couple of quarters, us make adjustments to that and highlight that to you. So annually, we're -- every quarter, we look at our prepay speeds, and annually, we do a big review. And what we saw, what we decided to do is true those speeds up. And importantly, speeds today are very much in line with what we see in the securitization market, how the market prices that risk as well, and so we're very much aligned. And so on average, what we tell you is that prepayment speeds across our private, our -- and our federal and consolidated loans have increased about 1 percentage point, and so think about the speeds now being in sort of the 5% to 6% range today.

Got it. And maybe to -- if you could, just to follow up on Lee's point but not looking backward, just, I guess, looking forward. I guess where I'm troubled is that the decisions that you're making in terms of the speed of these initiatives is what really does kind of cause the length of that suspension of the buyback. In other words, the stuff that happened historically has already happened. You're actually above the low end of your capital range. Your prepay speeds are increasing, so your portfolios are actually amortizing faster, your existing portfolios. And so I mean, when you think about where your tangible net asset ratio could be at midyear, I mean, could you kind of give us a sense as to where you think that might be by midyear 2018?

So I think, as you mentioned, there are a number of variables that come into play around where we're going to end up on tangible net asset ratio based upon the number of assets we originate, these prepayment speeds and a number of other things. I think importantly, clearly, what we did is we suspended the buyback program through 2018 to rebuild the ratio. I think what we clearly will always do is look at our capital position every quarter and make sure we're aligned with what our expectations are and where it is. And if we do find that we are putting ourselves in an overcapitalized position, and we clearly will have to take a look at that, I think what should give you some comfort -- and fully agree and understand the frustration, where the stock price is and the suspension of the buyback, but what should give you some comfort and what I've been impressed by is this company's ability to make sure that there's no -- the capital in the balance sheet is appropriately allocated, deployed back to shareholders so that we aren't in an overcapitalized position or there isn't any excess capital in the balance sheet. This is the way the company has been run since the spin and will be -- will continue to be run. And so I think, importantly for you, we will continually always assess, we will review. And if we do find ourselves in a position that we think that we have exceeded our goals or expectations, then we clearly will take another look.

But just to follow up on that, Chris. Because it actually is different than the spin -- since the spin because what's different now is you've made a series of decisions in terms of the -- number one, buying Earnest; number two, setting an origination goal for Earnest that's higher than its historical; and then looking out to starting to originate private loans. All of those 3 are different than -- since the spin, and each of them requires incremental capital. So my question is just simply, if you were to dial those back a little bit at a point in time when the stock is selling at roughly half of your expected intrinsic value, how do you think about whether that makes sense or not? That's the question.

I totally -- again, we understand your point. And I think our view is broadly that the capital is being retained within the company. Theoretically, the intrinsic value of the company should only increase. It's not going down or changing, it should only increase. And we do expect to fully return to a capital allocation program for shareholders' benefits and buybacks in 2019. And so inevitably, I understand the doubts you're talking about. There could be some variability about switching back or pulling back on originations. But as Jack mentioned, our belief is that by growing this franchise, building these assets which will produce very attractive returns in the future for shareholders, the value of that is very attractive, and we believe today that, that is the right approach. But again, as I said, we will always assess daily, monthly, quarterly what our capital position is, the return profile of what we're doing, and if we -- and if that equation changes, then we clearly could always change that.

Right. I guess -- and I don't want to beat this to death too much, but I'm also kind of thinking that you probably should be thinking about the forward plans for the asset side of the balance sheet given the capital consumption you talked about in terms of these loans. They have wonderful returns, but those returns aren't really realized until the second or third years of their life. And so there's a lot to be said for the timing of this versus the timing of where the stock is. It doesn't really need a response. I just wanted to say that.

We've taken it into consideration, and appreciate your feedback and the discussions we've had.

(Operator Instructions) Your next question comes from the line of John Hecht with Jefferies.

First question, I'm wondering if you guys can give us an update on the legacy bank portfolio acquisition opportunities. And I guess I'm asking that partially in the context of your capital allocation plan and any limitations that may come out of that with the focus on originating with Earnest.

Sure. So we've been, as we have through the -- since the separation, been very aggressive at buying portfolios of FFELP and private loans from sellers, principally banks. We've acquired nearly $30 billion since separation. And we expect the -- with the refi portfolios that we will add to our balance sheet, this year, we'll do over $10 billion of loan additions in 2017. The decision of banks to sell is impacted or influenced by various factors. We've been successful at acquiring virtually all of the large portfolios that had been offered for sale. We continue to pursue portfolios that are outstanding, and we think we're -- given our position of funding capability and, more importantly, how the portfolios are managed post purchase, puts us in the best position to acquire further loans, and we'll be aggressive at pursuing those.

And so do you see -- based on pipeline and discussions and this and that, do you see similar opportunity set over the next 12 months as you've had in the last 12 months?

It does -- they do vary. And the bigger portfolios take longer to negotiate and address, mostly on the decision-making of whether to sell or hold, so they're less even. There is some portfolio purchase activity that is very consistent year after year, and that's where we'll continue to participate. We'll give updates as to what we expect 2018 volume to be in January.

Okay. And then second question is can you just remind us -- I know there's a catch-up of NCOs for the Chase portfolio -- or the JPMorgan Chase portfolio you bought in Q4. Can you remind us what that catch-up is and what you expect the kind of run rate to look like in the subsequent quarters through next year?

So we had previously provided some guidance around what we thought the fourth quarter impact would be from defaults on the company. That number that was suggested was a $30 million number. Expectations thus far is we are on track that we'll exceed that numbers, so we feel very good about that estimate.

Exceed, meaning beat it?

Yes, beat it not worse.

Got that. And so will that be -- should we think then of a linear kind of correlation between that and the provision? Or will there be any change to the allowance ratio we should think about in the context of running those charge-offs through as well?

Well, you should think about linear going forward, and obviously, the portfolio amortizes down. And we will provide specific guidance in the fourth quarter, which should give you -- and 2018 view on what we think it looks like going forward. But like I said, we feel very confident in the number that we produced, and we will -- we expect to beat it.

Credit performance though is definitely improving across the board. And as we've demonstrated consistently when purchasing portfolios in the past, is our -- the way we approach customer contacts and deliver solutions that allow borrowers to successfully manage their loan has a material -- materially positive impact at reducing defaults. So we would certainly expect to be able to generate improvements over historical performance in any portfolio we acquire.

Your next question comes from the line of Mark Hammond of Bank of America.

I have 3 questions. So using the expected $1 billion in education refinanced loans expected to be originated at Earnest this year, you mentioned 5% equity capital. Would you provide what the remaining mix you'd expect to be between the debt, so like the ABS and unsecured, that would make up that remaining 95%?

Yes. So if you look at ABS securitizations today, and obviously, structures are different, but assume sort of 88%, 88% to 92% advance rate, and then you can then back into what that 5% equity component and then what other funding would take the remainder of the capital structure. So that should give you a sense. You can actually look at the recent securitization deal we did and look at the terms of it. It should be very akin to what happens going forward in the refi space. You could look at SoFi's recent deal, which priced right before our deal, so you can get a pretty good sense of the required capital to fund those transactions.

All right. That 85% advance rate, that seems a bit higher than typical.

I said 88% to sort of 92%.

That seems typical for a FFELP deal. So that would be on the private side, right?

Yes. You should be -- it should probably be helpful -- and we could probably send you the docs. I mean it's all public. But if you look at the SoFi and our securitization, you can see some of the details around advance rates, et cetera.

All right. And then in your diligence for Earnest, did you come across or ever think about what the potential size of outstanding loans today would make up that refinanced market opportunity?

Well, we think the total activity in 2017 is somewhere between $8 billion to $10 billion of refi originations. And we think our -- again, going back to our experience and data and the technology and tools that the Earnest team has built, it's really a powerful combination that is unique in the industry so that we would expect to be a dominant player in that space. And the beauty of that -- this particular market is those loans are generated each and every -- new loans are generated each and every year. And as borrowers graduate, obtain the employment, demonstrate a payment track record, we're able to offer them interest rates that they have earned as a result of their hard work in education and career advancements.

All right. And then lastly, I have a strategic question for you. So you've been expanding in the noneducation business services area. So now you've purchased an education asset-generation business in Earnest. Have you considered and would you expand into noneducation consumer loan asset-generation or origination businesses?

Well, clearly, we believe the customer base that we own loans to or we'll be making loans to as part of the Earnest have other financial needs. Those certainly represent opportunities for growth, less likely to be balance sheet opportunities for us and more likely to be either make-and-sell types of opportunities or partnership type opportunities, but they're clearly attractive consumers.

Importantly, there clearly was some opportunity with Earnest in that they have capabilities today to originate personal unsecured loans than they do mortgages today. But it's all to sell -- it's all full agreements and so wouldn't be a balance sheet item, but we do think that there was clear optionality in that capability to expand on the value of that franchise.

All right. Great. And then lastly, I'd just say appreciate the attention to bondholders are regarding your capital plan.

But I think said better, we took a holistic view towards all of our financing partners and our equity holders, and we think they will -- our decisions will be to the betterment of all of our constituents going forward.

Your next question comes from the line of Henry Coffey with Whitebush (sic) [Wedbush].

Yes. Tough call. Going back to Earnest. Based on The Wall Street Journal, I don't know whether what they said was accurate or not, it looks like you acquired this operation at a discount to the amount of money they had put into it. I don't know whether you could comment on that or not. But secondly, we hear a lot of this stuff out of these fintech companies that have pretty high opinions of themselves. And when we really dig into it, they basically seem to be doing everything that a normal finance company would do. So there's not -- I mean, when you look at the Earnest, that "use of data", I mean, are they doing anything that Discover wouldn't be doing or that a credit card company wouldn't be doing? And was this simply an opportunity to buy a pool of technology and a servicing platform at essentially a discount to its cost to build? Or how did you view it?

So the purchase price reflected, I think, what we thought was a good value for the opportunity to build a business here, and that's the combination of the technology, the people and the franchise that has been developed to date. What prior investors were willing to invest, that is -- was not particularly relevant to our decisions, right? We think this is a great transaction for us and for the Earnest team as well. We're adding a capability here that did -- that was a bit of a challenge, which is the funding side of the equation. And it's an infrastructure benefit that, as I said in my comments, that the 1 plus 1 here is greater than 2. We both firmly believe that. I think the -- I forgot your second question here.

Are they doing anything different?

Oh, are they doing anything different? So yes, so we've spent -- we've been watching -- like most financial institutions, we've been watching what's going on in the fintech space, and I don't disagree with you with your observation that there is less tech in some of these entities than they'd like you to believe. And I think what we saw here was the opposite, that the technology applications that were being brought to bear in terms of systems and solutions that they were offering to customers and the use of the data to drive lead generation, credit assessment and pricing dynamics is what really was unique. And when you can partner that with our data of 40 years of experience in terms of how students perform in different stages of their life cycle and different economic environments, it really represents something that's pretty powerful. And yes, I do think it is different than what you're seeing at other companies. And that's really the opportunity, is to capture and value to take advantage of it.

I mean, if you look at the securitization data, which is a very, very narrow snapshot, you can't tell the difference between, so far, an Earnest and the 3 or 4 other platforms out there. I mean, it's high FICO, low-coupon loans to people that have jobs. Is Earnest -- underneath that cover, is Earnest targeting anything unique? And how does that translate financially, either in terms of customer acquisition cost or ultimate portfolio performance?

Sure. So there's a couple of things there. Certainly, customer acquisition cost is a big driver of value creation. And if you got better tools and technology to do that, as I said, we believe they do, and as a result, they're driving -- creating assets at a lower cost of acquisition than others. The other piece is that FICO Scores are the known metric that people look at when assessing credit. But there are other borrower attributes that make a difference in performance. And that is -- using those other metrics, you can better assess credit risk and perform -- and future performance. And even within 700 FICO Score borrowers, there are borrowers who will go bad. And if you can assess who those are through other metrics and avoid them or price them appropriately, that's a positive. It also works the other way. There are borrowers who -- because they are relatively early in their repayment streams, who will be very high-performing customers but may not yet enjoy the same FICO Score because FICO is looking backwards, not forwards. And again, we can use that data to help expand the universe of customers we serve and provide them with financing options at the rate they've earned, and that's the opportunity here.

Is Earnest the platform of the future for your in-school lending? Or is that going to be targeting a different audience?

It's definitely targeting a different audience. One is going to school and one is graduating.

No, I mean past that.

No, I think the platform -- the tools and technology and the data are -- the metrics may be different, but the approach and the value-creation opportunities are the same.

Your final question comes from the line of Mark DeVries of Barclays.

Yes. What, if any, OpEx can we expect to come out to partially offset some of the revenue lost under the USA Funds contract?

So we -- in our legacy-related businesses, which are part of the FFELP and private businesses, those are amortizing portfolios and amortizing services. And as part of our annual review process or even periodically through the year, we're assessing what our needs are and what our operating expenses should be in those areas. We've been very aggressive at managing our costs, our budgets to reflect the revenue that we generate in those areas and the services that we provide. I think one thing we've done particularly well is improve our automation and efficiency in these areas, and you can expect to see that continue in 2018. So costs should be coming out and efficiency should be going up.

Okay. But you can't be more specific on how much cost related to that revenue might come out?

I prefer to talk about that in January.

Okay, fair enough. And then, Jack, I was also hoping you could help us think about what a reinstituted buyback could look like in 2019. Should it be somewhat similar to what you've been doing recently the last -- the incremental capital that's consumed from the $1.5 billion-plus of consolidation opportunity and the in-school? Or might you take a slightly more conservative approach than that?

So our capital return is more about prospective needs versus what happened in the past. So we would certainly take into consideration volume flows and expectations of generation of assets and needs to -- and opportunities to reinvest capital, but that's really been the same process we've deployed in the past.

Okay, got it. And how much capital should be consumed by the in-school? You indicated 5% on the consolidation. Is it going to be more like 8% to 10% on the in-school loans?

That has been historically the right range. Depending on the -- there are different segments of the in-school market that have different types of risk profiles. So depending on where our focus is and the assets that we generate, we would take a look at that.

I would now like to hand the call back over to Joe Fisher for any closing remarks.

Thanks, Andrea. I'd like to thank everyone for joining us on today's call. If you have any other follow-up questions, feel free to contact me. This concludes today's call.

Thank you for your participation. This concludes today's call. You may now disconnect.