Eagle Point Credit Company (ECC) Earnings Call Transcript & Summary

Greetings, and welcome to Eagle Point Credit Company's Third Quarter 2020 Financial Results Conference Call. [Operator Instructions] As a reminder, this conference is being recorded. I would now like to turn the conference over to your host, Garrett Edson of ICR. Thank you. You may begin.

Thank you, Rob, and good morning. By now, everyone should have access to our earnings announcement and investor presentation, which was released prior to this call, which may also be found on our website at eaglepointcreditcompany.com. Before we begin our formal remarks, we need to remind everyone that the matters discussed on this call include forward-looking statements or projected financial information that involve risks and uncertainties that may cause the company's actual results to differ materially from those projected in such forward-looking statements project to financial information. For further information on factors that could impact the company and the statements and projections contained herein, please refer to the company's filings with the Securities and Exchange Commission. Each forward-looking statement and projection of financial information made during this call is based on information available to us as of the day of this call. We disclaim any obligation to update our forward-looking statements unless required by law. A replay of this call can be accessed for 30 days via the company's website eaglepointcreditcompany.com. Earlier today, we filed our third quarter 2020 financial statements and our third quarter investor presentation with the Securities and Exchange Commission. The financial statements and our third quarter investor presentation are also available within the Investor Relations section of the company's website. The financial statements can be found by following the financial statements and reports link and the investor presentation can be found by following the presentations and events link. I would now like to introduce Tom Majewski, Chief Executive Officer of Eagle Point Credit Company.

Thank you, Garrett, and welcome, everyone, to Eagle Point Credit Company's third quarter earnings call. If you haven't done so already, we invite you to download our investor presentation from our website, which provides additional information about the company, including information about our portfolio and the underlying corporate loan obligors. For today's call, I'll provide some high-level commentary on the third quarter and more recent events. Then I'll turn the call over to Ken, who will walk us through the third quarter financials. I'll then return to talk about the macro environment, our strategy and provide updates on some recent activity. We'll also open the call to your questions. During the third quarter, the U.S. economy continued its gradual recovery as many shutdown orders were relaxed and governments globally provided liquidity to the market. Worst-case scenarios for corporate credit that some predicted back in March have not materialized. Today, many research desks are actually reducing their near-term default projections. With that said, we were very mindful that we're not out of the woods just yet with respect to COVID as cases have been rising. Given our strong balance sheet and ample liquidity, we've been able to find attractive CLO opportunities, both in the secondary and primary markets. During the third quarter, we deployed a little over $27 million of net capital into new investments. We also converted 3 of our existing loan accumulation facilities into new CLOs. During the third quarter, the company received recurring cash flows from our portfolio of $16.7 million or about $0.53 per weighted average common share. Our net investment income was $0.29 per common share in the third quarter, well above our common distribution level. We did realize some losses, selling a few investments, and that brought our NII, net of realized losses to $0.23 per common share. Net of distributions paid to common stockholders, our NAV increased by 13% during the quarter. While we're certainly pleased with the upward NAV movement, we're mindful that it still has lagged other risk assets as CLO equity often does. Our NAV remains below where it started the year, and we believe there may be potential for continued upside in our NAV. Looking to October, we see some even better news. We received recurring cash flows from our portfolio of $25.5 million during the month of October. This is an increase of 50% over the prior quarter. Cash flows were pressured in the third quarter due to an adverse LIBOR mismatch when many CLO rates were set back in April of this year. That has corrected itself, and now CLOs are benefiting from a positive LIBOR basis, driven by 1% floors, which are becoming increasingly common on many loans today. We expect this positive basis to continue as long as LIBOR remains low, and that will benefit CLO equity. In addition, several investments in our portfolio that deferred payments in July resume making distributions in October as they came on sides with their OC Tests. Beyond the positive trends in our cash flows and earnings, we remain -- we continue to maintain a very solid balance sheet. We have no financing maturities due before October of 2026. We have no secured financing whatsoever, no repost file financing and no unfunded revolver commitments. So while we are optimistic that the economy will continue to rebound, we remain very well positioned, we believe, to whether setbacks should they occur. We have ample dry powder, a little over $12 million as of October 31, allowing us to continue to be on the offense when we see opportunities. We also declared common distributions of $0.08 per share per month for the first 3 months of 2021, keeping with the rate that we've been paying over the past few quarters. Earlier, I talked about positioning the company's balance sheet. I also want to highlight the value of the right side of our CLO equity portfolio's balance sheet. We continue to believe our portfolio could withstand a prolonged recession should it occur, and we believe it could thrive in it. This is not because we're blind to defaults, but because the value that can be created by reinvesting within our CLO structures. A key metric to evaluate our reinvestment optionality is how much reinvestment period we have left in our portfolio. At quarter end, our CLO Equities portfolio's weighted average remaining reinvestment period stood at 2.6 years. This allows our CLOs to continue to be on the offense in volatile markets. This measure was 2.7 years at the end of June and 2.9 years at the beginning of 2020. So despite the passage of 9 months through our proactive management of our portfolio, our portfolio's remaining reinvestment period decayed by less than half. After Ken's remarks, I'll take you through the current state of the corporate loan and CLO markets and share our outlook for the remainder of 2020. I'll now turn the call over to Ken.

Thanks, Tom. Let's discuss the third quarter in a bit more detail. For the third quarter of 2020, the company recorded net investment income, net of realized losses of approximately $7.2 million or $0.23 per common share. This compares to net investment income, net of realized losses of $0.28 per common share in the second quarter of 2020, and net investment income plus realized gains of $0.37 per common share in the third quarter of 2019. When the unrealized portfolio appreciation is included for the third quarter, the company recorded GAAP net income of approximately $44.5 million or $1.40 per common share. This compares to GAAP net income of $1.71 per common share in the second quarter of 2020 and a GAAP net loss of $1.59 per common share in the third quarter of 2019. Just a reminder, our short-term cash flow generation is largely unaffected by the unrealized appreciation or depreciation we record at the end of each quarter. The company's third quarter GAAP net income was comprised of total investment income of $16 million and net unrealized mark-to-market gains of $37.3 million, partially offset by expenses of $7 million and realized losses of $1.8 million. As of October 31, the company had $12.9 million of cash on hand, net of pending settlements. As of September 30, the company's net asset value was approximately $268 million or $8.45 per common share. Management's unaudited estimate of the range of the company's NAV as of October 31 was between $8.53 and $8.63 per share. The company's asset coverage ratios at September 30 for preferred stock and debt calculated pursuant to Investment Company Act requirements were 288% and 433%, respectively. The current measures are comfortably above their statutory requirements of 200% and 300%. As of September 30, the company had debt and preferred securities outstanding totaling approximately 34.7% of the company's total assets less current liabilities, which is within, but at the upper end of our range of generally operating the company with leverage between 25% to 35% of total assets under normal market conditions. Moving on to our portfolio activity in the fourth quarter through October 31, the company received current cash flows on its investment portfolio of $25.5 million or $0.80 per common share. To highlight Tom's earlier point, this is a 50% increase compared to the $16.7 million or $0.53 per common share received during the full third quarter. Consistent with prior periods, please note some of our investments are expected to make payments later in the quarter. During the third quarter, we paid 3 monthly distributions of $0.08 per share of common stock and are paying the same amount for each month in the fourth quarter. Additionally, this past Friday, we declared common distributions for the first 3 months of 2021 in the same amount. In terms of our aftermarket offering program in the third quarter, the company issued approximately 82,000 shares of its common stock at a premium to NAV for total net proceeds to the company of approximately $0.6 million. I will now hand the call back over to Tom.

Great. Thanks, Ken. Now let me take you through where the macro loan and CLO markets currently stand, and I'll touch on our recent portfolio activity as well. Now the Credit Suisse Leveraged Loan Index continued to see a nice recovery, generating a total return of nearly 4% for the third quarter of 2020. As of September 30, the CS Leveraged Loan Index was down less than 1%. And as of November 13, the index had actually reached positive territory for the year. This is remarkable in our opinion, when one considers how dire the outlook was for credit back in March. According to S&P, 13% of the loan market was trading below 90% at the end of September, and that compares to 22% at the end of June. This is a very big improvement, and we're happy to see it. Despite the improvement in the loan market, we believe opportunities continue to exist for our CLOs to reinvest and build par through buying loans at discounted prices. Our CLO equity investments should be well positioned to continue reinvesting given the benefit of the long-term, locked-in-place, non-mark-to-market financing inherent in our CLO structures. Retail fund, loan fund outflows lightened a bit but still persisted in the third quarter, and we saw outflows of a little over $4 billion. On a look-through basis, the weighted average spread in our portfolio increased slightly from 3.55% in June to 3.59% at the end of September. An increasing number of loans in the market now include LIBOR floors, which further increased our interest collection when short-term rates are as low as they are at present. Simply, we're earning floored LIBOR, and we're typically paying LIBOR that is not floored or subject to a floor of 0. That's very, very powerful. The trailing 12-month default rate at the end of September stood at 4.17% according to S&P. While we still expect additional defaults in the coming months, the rapid recovery in the market is causing many research desks to reduce their projections of default rates, and this is a very hopeful sign compared to where things stood 6 to 8 months ago. While defaults will likely to continue -- will likely continue at elevated levels in the short term, we believe the corporate default rate will remain lower than it otherwise would have been had more loans featured financial maintenance covenants. The company's default exposure as of September 30 stood at 2.03%, well below the trailing 12-month default rate. Further, only 5.4% of loans in our portfolio mature prior to 2023, providing a significant majority of our corporate borrowers with years and years of runway before their debts are actually due. Our portfolio's weighted average junior OC cushion was 1.12% as of September 30, and that's up from 0.83% or 83 basis points at the end of June. That's a big improvement, but is still where -- it's still below where it stood prior to the COVID onset. This principally reflects the impact of 40% of all corporate loans being downgraded by the rating agencies. However, as previously mentioned, the pace of downgrades has slowed, and we've actually started to see some upgrades of stronger issuers. In addition, many of our largest holdings have significantly greater OC cushion than the average. By market value, 95% of our CLO equity positions that were scheduled to make payments in October did so. Think about that for a minute, 8 months after one of the largest crises we've seen in several years, 95% of CLO equity in our portfolio is paying current distributions. It's the primary reason why CLOs are such an attractive and resilient asset class and what attracts us to invest the company's capital and our own personal capital in the stock of the company. To sum up, cash flows on our portfolio increased by 50% in October versus the third quarter. Our balance sheet remains strong, and we have no debt maturities for nearly six years. We have $12.9 million of dry powder, ready to invest, are the long-term, locked-in-place, non-mark-to-market financing and our CLOs is significant, and we consider it to be a generally underappreciated advantage of CLO equity. And we believe our adviser's deep expertise and strong track record are keys to our success, particularly during uncertain times. We continue to closely manage our investment portfolio and remain opportunistic with respect to deploying capital. As a reminder, we know how CLOs have performed historically. Many consider 2006 and '07 CLOs to be some of the best vintages of the 1.0 era. If today, CLOs perform half as well as the 1.0 set did, we believe that this will be a very attractive ultimate outcome for investors. With that, we thank you for your time and interest in Eagle Point. Ken and I will now open the call to your questions.

[Operator Instructions] Our first question comes from Randy Binner with B. Riley.

Yes, I have a few. Just on the commentary on the OC Test. You mentioned it's trending better. And I think from Page 30 of your deck, it's up to 112 basis points. Yes, just as a reminder, is that the right figure to be looking at? And where could that trend back to, do you think? Can it get back to historic levels?

Sure. A very good question. Yes. On Page 30, the 112 basis points, that's the number. And you can see that's up from 83 basis points at the prior quarter end. Some of the drivers of that are going to be a factor of how much CCC we have in any given portfolio. And if you look back on Pages 25 and 26, you can see the CCC percentages or CCC+, Caa1 or lower, over on the right-hand side, I mean, you can see quite a few of the CLOs are over are over 7.5%. And there's a haircut taken in typical CLO. There's a haircut taken in the OC Test for excess CCCs. So as we -- the way to think of that this broadly is, to the extent we had more CCCs or the price of CCC loans fell, that would hurt the OC cushion. At the same time, if we have fewer and fewer CCCs or the price of CCC's rally, because the OC test looks at a market value of those -- of the CCCs, that we could actually get more cushion in the test. In addition, the other factors in there, sort of loan defaults would be a detriment to the extent there are trading gains or losses in the portfolio that can help or hurt the OC cushion. As to where it could get to, there's no functional limit on the upside. And to the extent you look at our average is about 13.5%, if you think of the average CCC as being around $0.70 on the dollar, just pick that as a number, you're taking a haircut over 7.5%. So 13.5%, minus 7.5% is 6 points. If times 30%, you could be -- we could be taking off 180, 200 basis points off the OC Test, on average, as a result of the high CCCs. So to your question, if the CCC buckets continue to come down, which is certainly the trend we've been seeing across many CLOs, you could see a path to get back to the 3s.

And then I guess the follow-up question is, understand the marks have been better and -- but the -- we have -- some of your credits -- this is my perception. And if I'm mistaken, please correct me, but I feel like some of your credits would be more exposed to kind of worsening COVID trends in the economy maybe versus corporate credit more broadly. And so is -- do you have a concern that it -- kind of the incidents of COVID continues to increase, there's more lockdowns, that you could see a reversal? Or do you feel like kind of the current credit trend is still catching up and moving in a positive direction? And if it's helpful to talk about certain subsectors to describe that, please do.

Sure. Sure. So to the first part of the question, I don't think we have a particularly adverse selection of above-average COVID exposure versus broader below investment-grade credit at a high level. And if anything, I think the trend in the portfolios have been for collateral managers to reduce exposure in COVID-related industries. So I struggle to see us being in an adverse selection, and perhaps, we've been fortunate enough to get to a better situation. So when we look at these shutdowns, and you look at changes in unemployment and changes in GDP, a year ago, people would be fussing over 50 basis point shift in GDP quarter-over-quarter. Then we have this gigantic drop in the second quarter and then the slow recovery now. And all of a sudden, a double-digit default rate from what was functionally near full employment at the turn of the year, yet we're sitting here with 95% of our investments still paying current. So at a high level, that's bad inputs and a good result. To the extent we see more shutdowns and/or more restrictions which is unambiguously a trend throughout the country right now, that's a bad fact, but I don't see it as a dire fact in that things that are helping us -- certainly, the Fed is providing a lot of capital into the investment-grade market, which then by sort of definition, brings along the high-yield market, which brings along the loan market as well. And frankly, the high-yield market has been a source of funding for many loan borrowers who might have needed additional liquidity. And whereas when we start this COVID period back in March, you would think broadly of the average leverage borrowers having 3 to 9 months of liquidity as a broad generalization. Hard to see companies not having doubled or tripled their liquidity runway through different measures of cutting expenses and putting additional debt on their balance sheets. While that debt does eventually need to be repaid, it's typically many, many years out in the future before that's due. So if we were to see another turn down and slowdown in economic activity, I think companies have much stronger balance sheets today, even the cruise lines and airlines and hotels -- Hilton just did a bond deal yesterday at 4%. So even guys who would be squarely in the crosshairs have raised a ton of liquidity, which is great, which would give them more runway. And then assuming we get some degree of political resolution in Washington, obviously, no guarantees there, it's hard not to see some degree of additional stimulus coming out, helping those who are unemployed. And when you think about how we made it through such a high unemployment rate with a relatively low default rate, that's largely because the government masked it, in my opinion, by sending checks to lots and lots of people every single week. Those checks have stopped. But to the extent something resumes in the new year, I think that would also be an offsetting fact. Long answer, but hopefully, you kind of get the flavor how we're thinking about it.

Our next question comes from Chris Kotowski with Oppenheimer.

I guess, my first question was, I mean, in terms of the distributions, you highlighted the fact that third quarter distributions are up 50% from the second quarter. And I'm wondering, is that a function of the improvement in the OC cushion? Or is it kind of the natural lumpiness from quarter-to-quarter? And is there a way to kind of triangulate between, say, moving back to a 3-ish percent OC average, OC cushion? What kind of impact that would have on your distributions from the CLOs?

Sure. A good question. A little bit of none of the above on the first part, and then we'll talk about how the OC Test could impact the second impact future distributions. The principal thing -- and if you look at the chart on -- in our deck on -- let me find the page, on Page 23. This is a chart we don't like. We started at $1.08, $0.90, $0.68, $0.53. That's a CAGR no one is a fan of. The Q4 cash flows in October as we shared, up over 50% from the Q3 levels. So what caused the degradation? Broadly, the drop in the cash flows from Q1 to Q2 were new OC failures and CLOs that were making payments, all of a sudden, diverting cash flows to either buy new collateral or pay down the AAAs. The drop from Q2 to Q3 broadly was due to a pretty grievous LIBOR mismatch back in the second quarter, in that CLOs -- each given CLO sets its LIBOR rate 4 different times a year. Loans set their LIBOR rates at -- probably one loan sets their rate every single day in a given CLO portfolio or pretty darn close to it. As LIBOR fell precipitously during the second quarter, many CLOs were burdened with paying roughly 1% LIBOR during the second quarter, but loans kept resetting lower and lower and lower. And in many cases, we actually had a negative LIBOR basis, which is what manifested itself in the reduction in cash flows from Q2 to Q3. So just due to a timing of the rapid movement in LIBOR rates, we were borrowing at 1% LIBOR proverbially and investing at a much lower LIBOR rate, which is bad. Roll the clock to October, 2 good things happened. Principally, LIBOR stayed low, which then -- and as more and more loans now have LIBOR floors, we were borrowing from the CLO debt investors at about 25 basis points LIBOR, but investing about half of the portfolio, a little less than half in loans with 1% LIBOR floors. So all of a sudden, whereas in the second quarter, we had a negative basis in LIBOR, here, we had a positive basis. To the extent LIBOR stays low, which is certainly the broad expectation in the market even the -- give you the current 5-year swap rate, is 45 basis points, current 3-month LIBOR is 22 basis points. So the market is saying that the short-term rates are going to stay low for quite a period of time. Many, many loans have 1% LIBOR floor. So we're actually now on a positive basis, which we expect will continue. And certainly, as long as rates stay low, that will keep going. So we don't think it's a one-hit wonder with the October payments being so far up. Then to the second part of your question, the OC cushion, the OC Test is a binary test. You either pass it or you fail it. So there's not a reduction, "oh, if the OC goes a little down, you trim a little." It's kind of an all or none test. You could partially pass it and cure, but that usually, when these things go, they go buy enough of a margin. I guess it's possible you could have a tweener. But nearly all of the CLOs, you can see are either comfortably passing or failing by a nontrivial amount. So that OC cushion, the real thing to look at is deal-specific. And so the question to ask yourself is, do I think these deals can come back on sides. Some of our highest payers are the ones that are off sides. Those are the ones often with the most aggressive and most spread-y portfolios. As we construct our portfolio of CLO equity, we seek to have both conservative portfolios and more aggressive portfolios, and it's a continuum. And as you'd expect in choppier times, the more aggressive deals frankly, are the ones that are more likely to be off sides. Principally concentrated in Marathon and Zais, if you had to kind of put it into 2 shelves. That said, each of the teams at those shops are working keenly. They have, in many cases, meaningful personal co-investment and even part of their fees are deferred. So there's significant incentive to get those transactions on side. I think some have the potential to, while others might have a more challenged period ahead of them. But so when you're looking at our future cash flows, the question is deal-by-deal. And broadly, we're seeing a trend of OC cushions improving on the failing deals, but it takes a while for some of the most severe portfolios.

Okay. And then you mentioned -- you talked about the LIBOR floors. And I've been kind of under the impression that most leveraged loans had LIBOR floors for years. And I'm just kind of thinking, is it the percentage different now than it was, say, in 2015 and '16 when we were also near 0 LIBOR?

Yes. So it -- so mindful that loans typically prepay -- so yes, is the short answer. Mindful that loans typically prepay at a pretty rapid rate, kind of 30%, 35% per annum in normal market conditions, it slowed somewhat during the COVID period. If you see on Page 19, we show the annual prepayment rate for loans, which I guess averages just less than 33%. So the universe of loans keeps changing. From 2015, indeed, the vast majority of loans had LIBOR floors as LIBOR crept up to 2%, 3%, it seems like so long ago, what do you know, those loan borrowers as they were issuing new loans or refinancing, they said, "Hey, maybe let's not put a floor." And the market kind of said, "We're at 3%, 2%, it doesn't really matter. That's fine." So now we're in a situation where probably between 40% and 50% of the market broadly has LIBOR floors. That said, the vast majority of new loans is getting issued today do -- or a significant majority do have LIBOR floors of new loans getting created. So the market kind of let those slip away in a high LIBOR day, but the trend is certainly coming back our way.

Our next question is from Mickey Schleien with Ladenburg Thalmann.

Tom, could -- how would you describe trading and CLO equity currently in terms -- in the secondary markets in terms of volume and bid-ask spreads? I'm asking because sort of in the middle of the year, it was very choppy with wide spreads, and that causes difficulty, both in terms of making investments and in terms of valuation. So curious where we stand now.

Certainly, volume has continued to pick up broadly. One of the things -- so it's the -- and we've been active on a number of investments. We both bought and sold in the secondary market for CLO equity and debt. So the market is open and active. You see a typical pickup in activity right about now, kind of right after payment date. So if the CLOs typically paid October 15, October 30, now is the time, oh, we just got the payment, people will look to refresh or buy or sell things in their portfolio. So we're probably at a period of time with some meaningfully increased activity. There's a number of BWICs today that actually have CLO equity on them, including maybe even 1 or 2 matchers within the Eagle Point complex. The bid-ask spread is an interesting dynamic in the CLO market in that the -- while some dealers maintain inventory and a significant block of inventory, the vast majority of CLO equity trades on a customer-to-customer basis with a dealer intermediating for a small fee, in the case of CLO equity, typically 0.25 point. So there might be bank A in between of customers 1 and 2. In theory, you don't know who the other person is and the dealer just takes a small markup. So it's not as if you're kind of working off an offer sheet, which you might be in the loan or high-yield market where the dealers are going to have between a half- and 2-point market. Here, you're able to get to a much tighter -- the difference between the buyers and sellers pay is much lower because you don't have a dealer taking risk in between. Against that, buyer and seller expectations may vary, and that does give rise to a nontrivial spread where you put a bond out for auction, the highest bid might be 60, you get a couple of bids in the high 50s, sell it below 65, so there's nothing to do and the auction fails or it's DNT-ed, is what it's called, does not trade. There's probably, I think, buyers have gotten a little more realistic and sellers have gotten a little more realistic. Just probably in April, that was close to no secondary activity. And I'll say, the volume has picked up slowly gradually. Not every month has been up month-over-month, but in general, the volume has picked up. While there's still a few unrealistic sellers out there. When you put stuff out for bid, you'll definitely get 5 to 15 bids, depending on the security. And it's really just your choice as the seller, if you want to take them.

That's helpful, Tom. And how would you characterize the investment opportunity in the primary markets within CLO equity?

I'd say it's recently gotten better in that probably the best piece of news is many CLOs are able to get done now with 5-year reinvestment periods again. To kind of roll the clock back, if we were talking in February, we would have said, oh, CLOs have a 5-year reinvestment period, and that's standard to your non-call. Then all the bad things happened in March, and by April, we were seeing CLOs get done on a static pool basis or with a 1-year reinvestment period. We didn't participate in any of those. We did begin to participate in CLOs that had 3-year reinvestment periods, which was kind of commonplace, say, between March and September. That's actually kind of -- those had 1-year non-calls, and that's actually going to help us because we did 2 -- maybe 2 CLOs back in late Q2, early Q3. We did an Octagon deal and a CIFC deal. Both of which have AAAs, like 150s, 160s, 170s context, which could easily be refinanced tighter. And thankfully, those only have 1-year non-calls. So while we had to suffer only getting a 3-year reinvestment period, you can be assured if markets stay the way they are today, we'd keenly try and refinance those deals as soon as next -- Q2 next year, whenever they roll-off of lock. Seeing deals go out to 5 years now is even better. But at the price of a 2-year non-call. But we are seeing AAAs broadly in the 130 context, which is generally in line with probably the long-term -- or the average over the last 7-plus years in the CLO market. So we're pleased to see...

It's pretty amazing when you think about it, right?

Yes. Well, I mean, again, remember where we were. I mean, 95% of our portfolio by market value is still making current payments. We were at a double-digit unemployment rate. You have radical double-digit GDP contraction. No CLO forcibly liquidated, some turned off payments for a while, and now they're gradually coming back on sides. We have a stable hand and ECC itself has a stable hand that. Obviously, we like everything to pay all the time, something bad is going to happen at some point in the future as well, but we've got the steady hand both within the CLOs and within the company to be able to weather those storms. Now that we're back to 5-year issuance, we'll probably turn up the gas a little bit. The one drawback when I talked about the price of loans, that so many -- what was the stat? Only -- it was pretty small, only 13% of the loan market was below 90. That is both good and bad news. It's improved from 22%, below 90 at the end of June, but when you're creating a new CLO, you want to buy loans cheap, when you've already got a CLO, you like loans to go up, I guess. So it -- the price of loans certainly has continued to rally. The loan index is positive for the year. So loans are not being given away by any stretch, making it a little more difficult to make new CLOs work. That said, deals are getting done. We have a number in formation phase. Whether we do anything more this year, kind of still to be determined, but we're actively evaluating.

Okay. Understand. Tom, if we look at your portfolio, broadly speaking, could you give us a sense perhaps of what percentage of your portfolio is in vintages you believe may not survive the impact of the pandemic and will have to eventually be unwound? This may be -- I'm thinking maybe the '14, '15, '16 vintages with a lot of oil and gas and maybe more retail and then things like that?

Yes, that's a very astute question. The -- that '14 vintage certainly was -- and for all participants, the '14 vintage, particularly the middle of '14 had particularly high oil and gas exposure in those portfolios. The beginning of '14 didn't -- in the end of '14, didn't -- generally, but that the second and third quarter really were -- was pretty darn high. So that's a bad fact. And many -- if you had started with high energy, you already kind of used some of your cushion dealing with that, when that all went pear-shaped in '15 and '16. When you look at our portfolio, probably the biggest thing I would look at is those that could be the most problematic are those with low or negative OC cushion and those that have nothing left in the reinvestment period. That's where you could see things be kind of the least good. And I'll even compare 2 CLOs here, let me just make sure I get the right row. If you look at Zais 3 versus Zais 5, so these are both CLOs that are failing their OC Test right now, Zais 3, we were able to get a reset off in Q2 of '16. We did that via Goldman, I remember. That deal has got 2.8 years left to go. And so lots and lots of runway. Zais 5, we worked on doing a reset of that. We struggled to get the AAAs placed. And unfortunately, on that one, the reinvestment period is over. So despite the reverse order, because 3 was reset, if I had a bet on one of those horses, I bet on 3 versus 5, just by virtue of it's got a lot more runway to go. So it's a combination of OC cushion, and remaining reinvestment period that should kind of form the collage of how you think about which deals might be tougher.

I appreciate that. If I could switch gears maybe to yields. There was about a 60-, 70-basis point decline in the average effective yield, excluding non-call deals. So I was curious -- we don't have a lot of time to look at your presentation materials before the call, so I was curious if there was something specifically driving that. And if we think about the future, you've talked about very -- relatively speaking, very strong cash flows in October, which implies cash yields meaningfully higher than they've been in the last couple of quarters. And I'm wondering, with that trend, can we expect you to book more of that into income as through the effective yield just to reflect the dynamics in the market?

Yes, let me hit 1 or 2 points, and maybe Ken will comment a little further around it. So 1 thing that moved against us, whenever when we model CLOs, the short-term basis risk of LIBOR high versus LIBOR or the high LIBOR that we set the CLO debt at versus the low LIBOR we were earning it in Q2, is not something that the average industry model really factors in. People kind of assume LIBOR cancels each other out with the exception of LIBOR floors. So broadly, if you were to look at where our yields were and what our cash flow projections were back at the beginning of the year, we wouldn't have projected the significant drop on a deal -- forgetting about deals that went off sides of deals that continued paying, we didn't model the drop in cash flows simply because we didn't model that LIBOR would be paying 1% and investing at lower than 1%. So when you have less cash coming in versus your accrual, then your kind of -- your amortized cost stays higher than it otherwise would have. So when you're going into recasting, and we recast these every quarter now, now we're starting in a situation where we have a higher-than-expected amortized cost. So whatever future cash flows you have are going to be -- it's just going to be a lower yield because you've got more basis that you're blending those cash flows over. And then, Ken, do you want to comment any further around that?

That's exactly right, Tom. It's a function of cash relieving your amortized cost as well as any accretive costs, which would include previous interest receivable, which was not fully relieved when cash flows came in lower than expected. So I think if you follow the cash, Mickey, that will be decreasing the basis where -- as long as we protect the cash flows in the future are continue to sustain, and in some cases, go up, the yield should follow. There may be a little bit of a lag because when those cash flows -- a little bit of even quarter lag. When those cash flows are coming in, they are amortizing the current cost, which is going to be used as a basis for the projection in the next quarter.

I understand, Ken. And my last question, Tom, triangulating your GAAP cash and taxable income is, obviously, very difficult for analysts or for investors. But...

And for a company management, just to be clear.

Yes, I know. Ken and I have talked about that. So this year or I guess, year-to-date, there's been some return of capital reported. I think some of that is obviously because in the first quarter, you -- at least on a GAAP basis, you over distributed but now you're under distributing. And then we've seen this dynamic where the CLO managers are reducing their CCC buckets. I suspect some of that is active selling and harvesting -- the consequence of that is maybe perhaps generating tax losses, which will depress taxable income. So looking ahead, I'm trying to understand where the distribution may end up when we're past COVID, given that NII is now running comfortably ahead of the dividend and things look like they're moving in the right direction.

Yes. The -- I think you hit the nail on the head. There's definitely some losses getting realized in the CLOs, even though they might not impact our current cash collections, can have the impact of sheltering taxable income. For some of the largest positions in our complex, which may include positions in ECC. We hired an accountant to do a couple of preliminary estimates. And I think the answer is, unfortunately, answers vary. Some CLOs, to the extent we own them in PFIC form are typically floored at flashing 0 income, they can't flash up a negative taxable income to us. But it wouldn't surprise me to see some CLOs in the portfolio that are still generating full cash flows without interruption, will be flashing 0 taxable income to us this year. So well, I think, while we like cash without the burden of tax, that's a good fact in general, against that, others will be failing OC test, but might generate -- not paying cash and actually still flash as taxable income because they didn't take the losses. So it will be a little more scattered. Overall, I think it's safe to say the taxable income will be down year-over-year. Against that, as you point out, we're -- the GAAP earnings are comfortably in excess of the -- the NII is comfortably in excess of the distribution, the cash is up 50 -- was in excess of the distribution last quarter, and it's up 50% quarter-over-quarter and expenses don't really go up that much, when cash goes up that much. So those are some good facts; hopefully, a path to both build back NAV over time, both through appreciation and getting extra cash off the investments. I'll say we -- when we set the new distribution rate back in the spring, we tried to be prudent and conservative with it. So we weren't trying to squeeze every last cent out. So perhaps there's more good things to come as well, but we'll continue to watch the portfolio behave. These are long-term decisions we make, not short-term decisions. And our number one objective, let's just keep the cash flow coming, keep hopefully getting more and more of the investments back on size, and that can kind of help the board set the distribution policy into next year.

[Operator Instructions] Our next question comes from Ryan Lynch with KBW.

Ken, it's been a really good discussion you guys have had so far this morning. I just have 2 questions. First one is probably for you, Ken. If I look at your net income this quarter of $1.40 per share and then I look at your $0.24 dividend, that's about $1.16 net income after distributions this quarter. Your NAV is only up about $1 this quarter. So I was just curious, could you bridge that $0.16 gap?

Sure. That would -- $0.16 would be the change in other comprehensive income during the quarter, which we report separately from the income statement. That $0.16...

What was driving that?