Golub Capital BDC Inc (GBDC) 2021 Q3 法說會逐字稿

Welcome to GBDC's August 10, 2021 Quarterly Earnings Conference Call. Before we begin, I would like to take a moment to remind our listeners that remarks made during this call may contain forward-looking statements within the meaning of the Private Securities Litigation Reform Act of 1995. Statements other than statements of historical facts made during this call may constitute forward-looking statements and are not guarantees of future performance or results and involve a number of risks and uncertainties.

Actual results may differ materially from those in the forward-looking statements as a result of a number of factors, including those described from time to time in GBDC's filings with the SEC. For materials the company intends to refer to on today's earnings conference call, please visit the Investor Resources tab on the homepage of the company's website, www.golubcapitalbdc.com and click on the Events/Presentations link. GBDC's earnings release is also available on the company's website in the Investor Resources section. As a reminder, this call is being recorded for replay purposes.

I will now turn the call over to David Golub, Chief Executive Officer of Golub Capital BDC.

Thanks, operator. Hello, everybody, and thanks for joining us today. I'm joined by Ross Teune, our Chief Financial Officer; Gregory Robbins, Senior Managing Director; and Jon Simmons, Managing Director. Yesterday afternoon, we issued our earnings press release for the quarter ended June 30, and we posted an earnings presentation on our website. We'll be referring to that presentation throughout the call today. For those of you who are new to GBDC, I want to briefly describe our investment strategy. Our investment strategy is, and since inception, it's been to focus on providing first lien senior secured loans to healthy, resilient middle market companies that are backed by strong partnership-oriented private equity sponsors.

The headline for the quarter ended June 30 is that GBDC's results were very strong. GBDC had solid net investment income, continued strong credit performance and robust new deal activity. We'll be discussing each of these topics in greater detail as we go through today's presentation. Gregory is going to start by providing a brief overview of GBDC's performance for the June 30 quarter and then he'll hand it off to Jon and Ross for a more detailed review of those results. I'll then provide some closing commentary, and then we'll open the line for questions.

With that, let's take a closer look at GBDC's results for the quarter and the key drivers of those results. Gregory?

Thank you, David. Turning to Slide 4 for the quarter ended June 30, GBDC's adjusted NII per share was $0.29. Adjusted EPS was $0.49 and ending NAV per share was $15.06. The key drivers of those strong results are summarized on Slide 6. Two key themes I want to highlight. First, our portfolio continued to perform well. Golub Capital Middle Market Report, or GCMMR, for June 30, which we published about a month ago, reflected this strength. The GCMMR normally looks at year-over-year revenue and profit growth of Golub Capital middle market borrowers, but our approach for June 30 was a bit different. We didn't think year-over-year comparisons were particularly informative as they would be comparing the lockdown economy of April and May 2020 to the booming economy of April and May 2021. We thought it would be more informative to compare 2021 to the pre-COVID economy of April and May 2019. And the results are striking.

The median EBITDA growth from 2019 to 2021 exceeded 30%. Strong earnings growth across GBDC's portfolio was reflected in the 4 positive credit quality trends listed on the right-hand side of the slide. We'll go into them in more detail shortly. Second, middle market new deal activity was strong. The quarter ended June 30 was a new record for Golub Capital origination volume. You may recall that the quarter ended December 31, 2020, was a record breaker at the time. Well, we beat that record by 30% in the June quarter. While the impact on GBDC's portfolio growth was muted by elevated repayments in the June quarter, we think the key takeaway is that GBDC is delivering strong originations amidst a robust M&A environment by leveraging the competitive advantages of the Golub Capital platform.

Let's now drill down on the 4 positive credit quality trends listed on the right-hand side of the slide, starting with our internal performance ratings. Slide 7 summarizes the positive trends in the internal performance ratings of GBDC's portfolio in the post-COVID period. Two key trends continued through the quarter ended June 30. First, portfolio companies performing materially below expectations in categories 1 and 2 remain very few in number. Those 2 categories constituted only 1.1% of the portfolio at fair value at quarter end. We've also seen continued upward migration in credit quality, a steady increase in categories 4 and 5, those are loans performing at or better than our expectations at underwriting, and a corresponding decrease in category 3. Those are loans that are performing or expected to perform below expectations.

Specifically, in the quarter ended June 30, categories 4 and 5 increased from 86.9% of the portfolio as of 3/31 to 89.4% of the portfolio as of June 30, an improvement of 13.4 percentage points year-over-year. Category 3 decreased from 12% of the portfolio as of 3/31 to 9.5% of the portfolio as of 6/30. Proportion of the portfolio rated 3 as of 6/30 was right in line with our pre-COVID normal of around 10%, as you can see from the fiscal year-end 2018 and 2019 data at the far left of the slide. A second key indicator of continued credit improvement is the fact that nonaccruals remain very low, just 1% of investments at fair value at quarter end. The nonaccrual rate was unchanged quarter-over-quarter, but about 50% lower year-over-year. We'll come back to this point in our usual discussion of GBDC's financial results.

Slide 8 shows 2 other indicators of improving credit quality, net realized gains and net unrealized gains. This slide provides a bridge from GBDC's $14.86 NAV per share as of 3/31 to its increased $15.06 NAV per share as of 6/30. Let's walk through the bridge. Adjusted NII per share was $0.29, in line with our quarterly dividend. No net realized losses were recorded during the quarter. In fact, there were $0.03 per share of net realized gains and net unrealized gains were $0.21 per share, excluding the purchase premium adjustment, reflecting the continued reversal of unrealized losses incurred in the March 2020 quarter. In fact, on a price basis, 90% of the unrealized losses recorded in the quarter ended March 31, 2020, have been recovered as of June 30, 2021.

Finally, Slide 9 shows that the continued strength and quality of GBDC's portfolio enabled us to further optimize the company's debt capital structure post quarter end. On August 3, 2021, GBDC completed its third unsecured bond offering. The $350 million issuance of new unsecured notes mature in February 2027 and have a fixed interest rate of 2.05%, the lowest coupon ever achieved by a BDC at the time. Pro forma for the August offering, unsecured debt represents approximately 50% of GBDC's debt capital stack. You can see from the chart on the right-hand side of the slide that GBDC is expected to have no contractual debt maturities until 2024 and the vast majority of GBDC's funding matures in 2025 or later. In short, we believe GBDC's balance sheet is stronger and more flexible than ever.

Let's now take a closer look at our results for the quarter ended June 30. And for that, let me hand the call over to Jon Simmons to walk you through the results in more detail. Jon?

Thanks, Gregory. Slide 11 summarizes our results for the quarter ended June 30. You can see in the column on the right that adjusted NII per share was in line with our quarterly dividend and that our credit results remain strong as GBDC generated $0.20 a share of adjusted net realized and unrealized gains. As a result, our net asset value per share at June 30, 2021, increased to $15.06. On August 6, our board declared a quarterly distribution of $0.29 a share, payable on September 29, 2021, to stockholders of record as of September 8, 2021.

Turning to Slide 12. New investment commitments totaled $614.7 million for the quarter ended June 30. After factoring in total exits and sales of investments of $583.5 million as well as unrealized appreciation and other portfolio activity, total investments at fair value increased by 1% or $44.3 million during the quarter. As Gregory noted, originations this quarter were strong, while repayments were also elevated. As of June 30, 2021, we had $45.4 million of undrawn revolver commitments and $171.9 million of undrawn delayed draw term loan commitments. These unfunded commitments are relatively small in the context of GBDC's large balance sheet and strong liquidity position. As shown on the bottom of the table, the weighted average rate on new investments and spread over LIBOR on new floating rate investments each increased slightly quarter-over-quarter.

Slide 13 shows that GBDC's portfolio mix by investment type remained consistent quarter-over-quarter with one stop loans continuing to represent 80% of the portfolio. Slide 14 shows that GBDC's portfolio remained highly diversified by obligor with an average investment size of less than 40 basis points. As of June 30, 96% of our investment portfolio remained in first lien senior secured floating rate loans and defensively positioned in what we believe to be resilient industries.

Turning to Slide 15. This graph summarizes portfolio yields and net investment spreads for the quarter, focusing first on the light blue line. This line represents the income yield or the actual amount earned on our investments, including interest and fee income, but excluding the amortization of upfront origination fees and the GCIC purchase price premium. The income yield decreased by 10 basis points to 7.4% for the quarter ended June 30, 2021. The investment income yield or the dark blue line, which includes the amortization of fees and discounts, also decreased by 10 basis points to 7.9% during the quarter. Our weighted average cost of debt for the aqua blue line decreased by 20 basis points to 2.8%, primarily due to the early redemption of $165 million in higher-priced SBIC debentures in the prior quarter. Our net investment spread, the green line, which is the difference between the investment income yield and the weighted average cost of debt, increased by 10 basis points to 5.1%.

With that, I'll hand the call over to Ross to continue the discussion of our quarterly results. Ross?

Thanks, Jon. Looking to the next 2 slides, nonaccrual investments as a percentage of total debt investments at cost and fair value remained low and consistent quarter-over-quarter at 1.4% and 1%, respectively, as of June 30. During the quarter, the number of nonaccrual investments remained unchanged at 6 portfolio company investments. As Gregory discussed in his opening commentary, as a result of continued strong portfolio company performance, the percentage of investments rated 3 on our internal performance rating scale decreased to 9.5% of the portfolio at fair value as of June 30. As a reminder, independent valuation firms value at least 25% of our investments each quarter.

Slides 18 and 19 provide further details on our balance sheet and income statement as of and for the 3 months ended June 30. Turning to Slide 20. The graph on the top summarizes our quarterly returns on equity over the past 5 years, and the graph on the bottom summarizes our regular quarterly distributions as well as our special distributions over the same time period. Turning to Slide 21. This graph illustrates our long history of strong shareholder returns since our IPO. As illustrated, investors in GBDC's 2010 IPO have achieved a 10% IRR on NAV since inception.

Slide 22 summarizes liquidity and investment capacity as of June 30, which remains strong with over $800 million of capital available through cash, restricted cash and availability in our various credit facilities. We also highlight our continued progress in optimizing the right-hand side of the balance sheet. Three key highlights here. First, on April 13, 2021, we amended our revolving credit facility with Morgan Stanley to, among other things, extend the reinvestment period to April 12, 2024, from May 3, 2021; extend the maturity date to April 12, 2026, from May 1, 2024; and reduce the interest rate on borrowings to 1-month LIBOR plus 2.05% from 1-month LIBOR plus 2.45%.

Second, on July 16, we issued a notice of redemption to redeem all of the $189 million of notes issued under the 2020 debt securitization, which are priced at 3-month LIBOR plus 2.44%. This redemption is expected to occur on August 26, 2021. And third, as Gregory mentioned earlier, on August 3, we issued $350 million of unsecured notes, which bear a fixed interest rate of 2.05% and mature on February 15, 2027, bringing unsecured debt up to approximately 50% of GBDC's total funding mix.

Slide 23 summarizes the terms of our debt capital as of June 30. And lastly, Slide 24 summarizes our recent distributions to stockholders. And most recently, our Board declared a quarterly distribution of $0.29 per share payable on September 29, 2021, to stockholders of record as of September 8, 2021.

With that, I'll turn it over to David for his closing remarks. David?

Thanks, Ross. To sum up, GBDC had a very strong quarter. Adjusted NII matched our dividend, realized and unrealized gains were substantial and robust new origination enabled the portfolio to grow despite unusually high payoffs. Let me talk about our outlook, and then I'll take your questions. The headline is the same as last quarter. We're cautiously optimistic. I'll start with why we're cautious. We've been concerned about COVID variance for some time. You've heard us talk about this in our last 2 earnings calls. As much as we'd all like to put this tragic pandemic behind us, it seems we're far from done with hundreds of millions of COVID cases around the globe, 85% of people worldwide not yet fully vaccinated and the possibility of more mutations.

That said, there are also reasons for optimism. We believe GBDC is prepared for this environment that we're in right now and has a set of powerful tailwinds. I'll focus on 3. We've spoken about them before, but they bear repeating. The first tailwind is GBDC's strong portfolio performance. We've highlighted throughout today's presentation the positive credit trends since March 31, 2020. Our pre-COVID underwriting has proved to be strong. Realized and unrealized gains and losses for the 18 months from January 1, 2020, through June 30, 2021, have netted to a loss of only $7 million or 0.16% of the portfolio at cost. That's an annualized loss rate of less than 11 basis points. And as Gregory described, the portfolio today has very low nonaccruals in minimal category 1 and 2 loans. So it's apparent we're not going to be distracted by needing to play defense on a troubled portfolio.

A second tailwind is the attractive opportunity set before us. 2 of the last 3 quarters set new records for Golub Capital origination. All signs are pointing to a robust middle market M&A activity in the second half of the year. While it's too early to tell if we'll set another record before year-end, we think Golub Capital is capturing more than our fair share of attractive deals. The competitive advantages of leading lenders have grown stronger through this COVID period, advantages like scale and sponsor relationships, incumbencies, breadth of solutions, industry expertise and reputation for reliability. We believe GBDC has a compelling opportunity to grow its portfolio in this environment without compromising on credit quality.

And finally, a third tailwind. The third tailwind is that GBDC has ample liquidity and flexibility to capture opportunities. We've achieved our goal of substantially increasing the proportion of unsecured debt in GBDC's funding mix, while keeping our cost of debt very low. Unsecured debt is now about 50% of our debt stack, and we believe GBDC has among the lowest unsecured funding cost of any BDC. GBDC's debt stack is well diversified, long-term, low cost and highly flexible. We're currently operating at the low end of our target leverage range of 0.85 to 1.15x debt to equity, and we think we have room to operate closer to the high end of that range in the coming quarters, which would help drive even stronger earnings power for the company.

Thank you. Operator, please open the line for questions.

(Operator Instructions) Your first question comes from the line of Finian O'Shea with Wells Fargo Securities.

First question, David, on I guess the comments you were just touching on, on the market opportunity and your market share, which remains very strong. Can you go specifically into the comeback of these larger private $1 billion, $2 billion transactions? Last time around before COVID, you were the obvious ring leader in these -- especially on a senior basis. And this time around, there's a couple more big players in those, as I'm sure you're familiar with. So how do you feel that your market share and competitiveness in terms of underwriting is holding up? And then as a second part there, how much of a challenge is that? Is -- do you think that paper is much better than the core middle market paper, or is it less of a challenge to the Golub platform?

Sure. Thanks, Fin. Thanks for your question. So by way of context first, we -- at Golub Capital, we underwrite loans that range in size from $10 million or $20 million at the low end up to multiple billions. As you pointed out, Fin, we've been a market share leader in what we call mega one-stops, which are $500 million and up one stop since those really began to be on the scene in 2019. Mega one-stops remain a relatively small portion of the overall mix of what we do. The predominance of what we do is financing companies that generate between $20 million and $50 million of EBITDA. Those are not companies where a mega one stop would be appropriate. In the last quarter, we closed 95 separate loans representing a bit more than $9 billion in total commitments. So if you just do the math on that, you can quickly calculate that our average loan size is in the approximately $90 million range.

Over the course of the last couple of years, we've seen, and as you pointed out, Fin, we've been a pioneer in expanding the role of private debt in larger-sized transactions. And I've said previously that there are a couple of key drivers of that phenomenon. One is that sponsors increasingly are looking at buy-and-build transactions, transactions where they are building a company over time through acquisitions as a means of creating a company of great scale and of great promise. They're not just doing financial engineering anymore. They're looking at creating great companies in large measure through acquisition strategies.

The broadly syndicated bank loan market is not an extremely efficient way of financing a company that's doing serial acquisitions. What many sponsors have found is that using a one stop and a series of either delayed draw term loans or serial expansions of that one stop can be a more effective way of financing their portfolio company that's doing serial acquisitions. Until relatively recently, that large scalable one stop was not on the menu. It wasn't 1 of the choices that a sponsor could choose from if they were looking at financing a company that had loan needs in excess of $500 million. That's now changed. And as you point out, in the last 6 months or so, we've seen transactions as high as $3 billion being pursued by private market by direct lenders. And we're not alone in this. We're still among the market leaders, but there are a number of very large players who are also capable of playing in this arena.

My view is that the mega one stop product is a great option for sponsors. It is not always a great opportunity for direct lenders. We here -- just as we are in smaller loans, we need to be very selective. We need to make sure that we're backing really good companies, that the terms and conditions of the loans that the pricing of the loans are attractive relative to other options that we have. So we're always going to be evaluating the relative attractiveness of different niches within direct lending that we operate in. And we're purposely going to be moving around as we see the best opportunities arise in 1 area, 1 industry sector, 1 geography versus others. And I think that's what we've been doing over the course of recent months.

That's helpful. Just a follow-on on the equity co-investment. It looks like you have a fair amount this quarter, at least versus historical. And given you're able to earn your dividend at such a low rate of leverage, one might say that you have built an ability to swing the bat much more often on equity co-invest. So any thoughts on the above there.

We've been pretty consistent in our percentage of the portfolio that's in equity co-investments. If you look at Page 13 of our investor presentation, it's been in the 2% to 3% range for an extended period of time. I think that you can reasonably expect that we're going to continue to grow the portfolio, that we're going to move our debt-to-equity ratio from where we are now, which is at the low end of our range more toward the middle or higher end of our range. I don't anticipate a meaningful change at this point in the mix that we're going to see. Of course, that's always subject to change based on market conditions, but that's my expectation now.

Your next question comes from the line of Paul Johnson with KBW.

First, I know you guys were just slightly below the low end of your hurdle rate this quarter and, therefore, did not earn the incentive fee. I'm just curious, are you okay with, I guess, operating around that area kind of right at or even below the low end of your hurdle? Or would you -- is the goal to essentially generate an ROE that's maybe above that 8%? I'm just trying to get your thoughts around the hurdle rate.

Yes. I think this quarter was a bit of an anomaly because the degree of repayments was as high as it was. My expectation, as I mentioned, is that we're going to see growth in the size of the portfolio. That in turn will grow net investment income and will give us more pre-incentive fee net investment income. I think when we look backwards, we'll see this quarter as a bit of an anomaly in the respect that you're mentioning. I think it's good for shareholders if we can operate in or above the catch-up as opposed to below the catch-up provided we can do so without taking too much credit risk. Right now, I think we can do that.

Great. That's very helpful. And then just 1 on your software lending portfolio. It's, I think, 26% or so, I think, from the slide deck of your portfolio. I'm just maybe trying to get your thoughts on how you guys view that market today. How you guys view that sort of competitive landscape? Obviously, we've seen a lot of growth and popularity of that type of lending. So is there anything that you guys see differently today maybe versus several quarters ago, or what are you looking out for in the new deals that you evaluate there today?

Sure. Again, let me just provide some context. So we've been leaders in software lending to sponsor-backed companies for more than a decade. I think we have a larger portfolio and more transactions under our belt in this sector than any of our competitive brethren. It's an area where we have very strong sponsor relationships, very strong incumbencies because of the portfolio that we've built and very strong expertise. We have a group within our underwriting team that specializes in software lending. We think we're very good at it. Our results over time in software have been outstanding, and we think that it's a robust area for future opportunity.

I think your statement is fair that we're seeing somewhat more competition in the software area than we did years back. But the flip side is also true that the private equity ecosystem, the component of that ecosystem that software companies continues to grow. And so we continue to find very attractive opportunities in the software space. We really haven't changed our approach in any meaningful way. We continue to be focused on really high-quality companies with mission-critical software tools that have been well integrated and are difficult to rip out of their clients with high recurring revenue streams and high renewal rates, the same sorts of underwriting criteria that we started out investing in the industry with more than 10 years ago.

Got you. And then my last question was just again your thoughts around the market for one stop unitranche loans versus maybe the senior loan, first lien, traditional first lien, second lien structure. What we've kind of seen the returns have compressed over time, obviously, in the unitranche market. I'm wondering, it doesn't seem to show up in your new investment mix, but do you evaluate those 2 markets any differently today in terms of just the value proposition of one stop loans versus the first lien traditional structure?

So every time we're looking at a new transaction, we're thinking about what the right way to play in it is, and what the right answer is for the transaction. And we'll proceed with a first lien solution if we think that it's compelling from a risk-reward standpoint and right for the transaction, or alternatively will proceed with a one stop solution if we think the risk/reward is compelling, and it's right for the transaction. So it's a multifaceted test that we use to assess what's the right solution to be emphasizing in our discussions with sponsors. At the end of the day, obviously, it's -- the sponsors make that choice. We don't make the choice as to what financial structure to put in place.

But we do have choices about where we play, and where we don't play. Right now, I would tell you that we continue to find a lot of attractive one stop opportunities. To your point, we're seeing a lot of steadiness in our income yield and in our weighted average net investment spread. I think you are seeing meaningful compression in junior debt spreads. Second lien spreads, in particular, have come down. And so I think perhaps that's putting some pressure on -- if you think about a one step as being an instrument that's priced as a hybrid that could be seen as putting some pressure on one stop spreads as well. But I think the data suggests more steadiness, more continuity than change there.

Your next question comes from the line of Ray Cheesman with Anfield Capital.

David, as we approach the end of the year and we get closer to LIBOR going puff and going away, do you perceive there to be any challenges in rolling all of the clients over to a -- I think they're now talking about SOFR term as the preferred way they're going to steer everybody. Is everybody ready for that? Do you perceive there'll be any impact on any income lines in your P&L?

So thanks for the question about LIBOR and SOFR. I think you may know I serve on the Board of LSTA. The main industry trade association has been very involved in this LIBOR transition in ensuring from an industry standpoint that the industry is ready. We, at Golub Capital, have also dedicated significant resources to make sure we're ready. Look, I think it's going to be a meaningful amount of work. Whether this transition happens at the end of this year or later, I think that's still an open question. I'm confident that whenever the timing is, we'll be ready and we'll have the resources in place to do the work with our borrowers to make sure that whatever changes are required in loan agreements are made. This is going to be a very significant lift, but I'm not -- I say that from a work standpoint, not from a risk standpoint. I think from a risk standpoint, it's quite under control.

Super. Glad to hear. Based upon the experience that you've had, when loans -- at the beginning, you were saying that the number of loans outperforming expectations, category 4 and 5, has increased at a very good speed coming up out of the lousy period a year ago. When your portfolio performs above expectations, should we expect to see a higher churn, higher repayments than otherwise because, obviously, they've got higher profit levels?

Sure. Let me clarify 1 thing. Category 4 is performing at or above, and Category 5 is performing above. So the statement that I hope we made before, I'm not sure exactly how we phrase this, is that the proportion of our portfolio that's performing in categories 4 and 5, meaning they're performing at or above expectations, has grown significantly. If you flip to Page 17 of our presentation deck, you'll see that those 2 categories are just under 90% of our loans as of June 30, 2021. And that's back in the range of the pre-COVID numbers.

I think you're on to an important point, which is more relevant for category 5 loans than for category 4 loans. I think category 5 loans do have a tendency to be refinanced or repaid more quickly than loans in other categories. And I think that has in part of the story of the more rapid than expected repayment rate that we saw in this last quarter. I don't think that's the main factor. I think the main factor is the very rapid pace of M&A that we've seen in the middle market generally. But I think you make an interesting point, which is that the category 5 loans do tend to refinance more quickly than 1, 2, 3 and 4 loans.

It's always that you lose the good ones, right?

Yes. It's the nature of credit.

Last is kind of a -- it's an open-ended question, and you've done unbelievably well at steering the company through a credit thunderstorm. You've lowered the cost of your funding. What are we looking for in the next couple of quarters to move the whole organization back from? I'm just going to use the base number for shareholders, $0.29 to $0.32. Is it a function that we need the macro rate environment to change? Or is it something else?

Great question. So you'll recall in prior quarters, I've talked about how our dividend policy has historically been linked to our NAV. We seek to, over time, have a dividend that corresponds to about an 8% annual rate of our NAV per share, and we seek to change it rarely and to increase it over time. So the path for us to raise our dividend is twofold right now. It's to increase our NII per share, pre-incentive fee by increasing our portfolio size and to continue to see some net unrealized and realized gains that increase our NAV per share. And those 2 will put us in a position where we can start to raise the dividend from $0.29 to north of that.

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

Congratulations particularly on all the work you've done on the liability side and everything else as well. On the kind of a follow-up to these other questions, to get back into just the catch-up, no pun intended. What you need to do to oversimplify is grow the portfolio a little bit. And you've talked about you're comfortable with that going forward. I mean do you expect repayments from a really elevated level this quarter, obviously? So is that going to moderate somewhat? And I mean, obviously, I'm not particularly concerned about your ability to originate debt, but the repayment question is much harder to predict for everybody. What kinds of trends in there?

Let's look together at Page 12 of the presentation and the second line in the table, exits and sales of investments in corporate new payments. And you can see that the June 30, 2021, quarter at $583.5 million is the outlier on this chart. It's literally more than twice the December 31 quarter, 1.5x the March 31 quarter. So if you ask me to make a prediction, my prediction is that we will not see a sustained level of exits and sales of investments at the June 30 level. Over time, in my experience, our assets tend to have a weighted average life between 2.5 and 3 years. The $583.5 million is effectively a weighted average life that's closer to 2. So I just -- I don't see a high likelihood of that sustaining over time, Robert. I think we're going to be able to continue to make good progress on originations. I think exits and sales will moderate, and I think we'll start seeing some portfolio growth.

Excellent. And then 1 more, if I can, on the credit question, and this may be overly picky, but I'll ask anyway. Your rating 3 assets, right, slightly below expectations in terms of performance, 9.5%. It's improved sequentially again. It's still a tiny bit above where it was in 2019. I presume those are still the COVID-impacted assets that aren't in categories 1 and 2. Of the ones -- of the assets that are category 2 this quarter and were last quarter because, obviously, some of them got upgraded, what's the performance like for those assets? I mean, yes, they're below expectations, but are they less below expectations this quarter than they were last quarter, if that makes sense?

So a couple of comments. First, again, let's look at the numbers together. If you flip to Page 17 of the presentation, you're correct that category 3 assets at 9.5% now are a little bit above where they were at the end of 2019, but they're actually lower than they were at the end of 2018, 2017 and 2016. So I think it's fair to say that the overall picture of performance ratings right now looks very much like it looked in pre-COVID periods. As to the specifics of the category 3 loans, most of them were actually challenged loans pre-COVID. These are not so much COVID-impacted loans as they are just loans that have been underperforming for some time. Are we making progress and seeing improved performance from this group? In many cases, yes. I can't speak with specificity in this forum on a loan-by-loan basis. But we have, as you know, a workout team that focuses on working with management teams to improve our underperforming credits, and they do a really good job. So I'm cautiously optimistic that we're going to see continued improvements in the portfolio over the coming several quarters.

You have a follow-up question from the line of Ray Cheesman with Anfield Capital.

I wanted to specifically say thank you very much because you give us great comfort when you release that middle market report, and you show us things like 31.5% improvement in earnings in the portfolio performance. I'm wondering, it is already the 10th of August and your measurement period for the next report closes in about 20 days. Will we be as happy when we see the trend in the next report? Will the good stuff continue?

I think that's more speculative than I'm prepared to answer right now on August 10. But thank you for the compliment. And I appreciate your enthusiasm for the GCMMR. Phyllis, do we have any last question?

At this time, there are no further questions. I would like to turn the call back over to management for closing remarks.

Great. Thank you, Phyllis, for your help, and thank you to all of you who have been listening today. We very much appreciate your support. As always, if you have questions that we didn't get to today, please feel free to reach out to any of the 4 of us from GBDC management, and we look forward to talking to you next quarter.

Ladies and gentlemen, that does conclude today's conference. We thank you for participating. You may now disconnect.