Golub Capital BDC (GBDC) Q1 2023 Earnings Call Transcript

Golub Capital BDC (NASDAQ:GBDC) Q1 2023 Earnings Conference Call February 9, 2023 1:00 PM ET

Company Participants

David Golub – Chief Executive Officer

Greg Cashman – Head, Direct Lending Group

Matt Benton – Chief Operating Officer

Chris Ericson – Chief Financial Officer

Conference Call Participants

Jordan Wathen – Wells Fargo Securities

Robert Dodd – Raymond James

Ryan Lynch – KBW

Operator

Hello everyone and welcome to GBDC’s December 31st 2022 Quarterly Earnings Call. Before we begin, I’d 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 SEC filings.

For materials we intend to refer to on today’s earnings call, please visit the Investor Resources tab on the homepage of our website, which is www.golubcapitalbdc.com and click on the Events Presentations link. Our earnings release is also available on our website in the Investor Resources section. As a reminder this call is being recorded.

With that, I’m pleased to turn the call over to David Golub, Chief Executive Officer of GBDC.

David Golub

Hello everybody, and thanks for joining us today. I’m joined by Chris Ericson, our Chief Financial Officer; Matt Benton, our Chief Operating Officer; and Greg Cashman, who Heads Golub Capital’s Direct Lending Group.

For those of you who are new to GBDC, our investment strategy is and since inception, it has 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.

Yesterday, we issued our earnings press release for the quarter ended December 31, 2022 and we posted an earnings presentation on our website. We’ll be referring to this presentation during the call today. As with last quarter, we’re going to follow a new agenda for these calls and that new agenda has me leading off with headlines and with an overview.

Let me start with the headlines. GBDC’s performance for the quarter was solid and it was consistent with our discussion last quarter. We saw strong growth in net investment income and generally stable credit trends.

What has changed is our macro outlook. I’ve said for the last that we’re in a period of unusually high uncertainty. I think that uncertainty is now easing and we’re in, and I think we’re likely to stay in, a period of muddling growth. I’m going to elaborate later on what I mean by this, how we’re responding to it, and why we think muddling growth can be good for Golub Capital BDC.

After that, Chris and Matt are going to go through the financial statements for the quarter in detail. And finally, I’ll come back and make some closing remarks and take questions.

Before we jump in, I also want to mention that we intend to file our quarterly equity investor presentation over the next few weeks. You’ll recall last quarter, I mentioned that this is a new presentation for us that we plan on updating each quarter. We hope you’ll find it a useful source of additional information on GBDC and on Golub Capital.

Okay, let’s start with a summary of performance for the quarter. Adjusted NII per share increased by 12% to $0.37 from $0.33 per share in the quarter ended September 30th, 2022. This equates to an adjusted NII return on equity of 10%. The increase in adjusted NII per share was driven primarily by rising base rates and by higher spreads.

Now, you’ll recall from last quarter’s earnings call that we believe higher base rates and higher spreads have materially increased GBDC’s earnings power. This increase in GBDC’s earnings power led to our decision last quarter to increase GBDC’s quarterly dividend by 10% to $0.33 per share. We think GBDC’s record adjusted NII per share for the quarter ended 12/31/2022, and its dividend coverage ratio of 112%, we think those validate the decision to raise the dividend. I’ll come back to how we’re thinking about the dividend going forward in my closing remarks.

Credit trends in GBDC’s portfolio remained generally stable. Realized credit losses were low. In fact, GBDC had a net realized gain of $0.02 per share for the quarter as capital gains on equity co-investments more than offset realized losses. This is a pattern we’ve seen consistently since GBDC’s inception. We did see some incremental spread widening in the quarter and that drove some incremental unrealized losses. As expected, we also had a small number of borrowers show some deterioration in credit performance and this is reflected in a small uptick in non-accruals and a small uptick in the percentage of our borrowers in performance rating categories one, two and three. But the big picture is that credit performance remains quite encouraging, portfolio companies are generally healthy and growing, and I’ll give you more detail on this in a few minutes.

Overall, our view is that GBDC’s performance in fiscal Q1 was quite solid. Before Matt Benton and Chris Ericson walk you through the financial results, I want to shift focus and talk for a few moments about our outlook. It may seem a bit odd to talk about the future before we fully unpack the quarter that just ended, but there’s a logic for our approach. The reason is that our outlook has changed since our last earnings call and this change has implications for the key issues that we’re focused on and that we think investors should be focused on.

One of the key themes we talked about last year was our belief that we were in a period of unusual uncertainty. I talked about how there were an unusually high number of powerful vectors that were impacting the economy and that we’re moving in different directions – inflation, rising interest rates, volatile energy and other commodity prices, changing consumer behavior post-pandemic, the Ukraine war. And I said that these different vectors meant it was very difficult to predict where we were going. We thought there was an unusually wide range of plausible scenarios for the coming period.

In recent months I think the range of plausible scenarios has narrowed pretty considerably. There’s still a fair bit of uncertainty. But in our view the last few months of data show that inflation has already decelerated strikingly, shows that interest rates are near their likely peaks. Occupancy costs, energy and commodity prices, they’re all generally declining. The unemployment rate has stayed low and consumer behavior has changed far less than many feared. All this means the odds of a deep recession look much slimmer today compared to three to six months ago. And the odds have increased that we’ll be in a period of muddling growth, and likely in a period of muddling growth for a sustained period.

So what does this changed outlook imply for GBDC? In our view it gives us higher conviction about our answers to three critical questions that investors frequently ask us. The three questions: First, how is GBDC’s portfolio doing? Second, will we see a spike in credit losses in 2023 or 2024? And third, net-net, are we going to see more writedowns, or are we going to see reversals of some of the writedowns GBDC has already taken?

I want to drill down on each of these three questions. I’m going to take the first one. I’ll tag team the second with Greg Cashman and then Matt Benton is going to answer the third question.

So the first question, how is the portfolio doing? The portfolio is doing well. The Golub Capital Middle Market Report for calendar Q4, which we published a few weeks ago, showed a positive surprise. Median profit growth accelerated from calendar Q2 and calendar Q3 levels and exceeded inflation by a significant margin. This was sharply different from the prior two quarters when median profit growth was a lot lower. Median revenue growth stayed strong in calendar Q4, consistent with prior quarters. Both the revenue growth and the profit growth numbers exceeded our expectations and this was true across all four sectors we track: consumer, healthcare, industrials and technology. We think these strong results reflect that our borrowers are adapting ably to the headwinds that started in the spring.

Second question: will we see a spike in credit losses? I want to take this question in two parts. I’m going to describe what I think is a bad way to answer the question. And then I’m going to ask Greg Cashman to walk through what I think is a better way to answer the question.

The bad way to answer the question involves a shortcut. It’d be great if we could answer the question with some quick to prepare or easy to understand set of quantitative metrics. And in fact some people try.

I don’t mean to pick on some of our peers, but many of them are showing charts these days that look at average interest coverage ratios. Sometimes they also show what happens to these coverage ratios using different assumptions about rising interest rates or declining underlying company EBITDA. This kind of chart, they sound instructive. Let me walk you through why I think they’re not.

First, interest coverage is EBITDA divided by interest expense. Sounds straightforward. But what’s EBITDA? Are we talking about GAAP EBITDA, or credit agreement EBITDA, or adjusted EBITDA, or some other measure? Particularly in a period of rapid change and uncertainty like the one we’re in now, all these measures are flawed. None really address what we really want to measure, which is go-forward earnings power and go-forward capacity to generate free cash flow.

The denominator, interest expense, is also problematic. What’s your base period? Are you factoring in the forward curve? Are you factoring in hedges or caps that the borrower may have in place? Again, this approach doesn’t address what we really want to measure, which is go-forward interest expense net of hedges over the next several years.

Another problem is this analysis assumes ceteris paribus. It assumes all else equal. But all else is rarely equal. What we really want to know is whether the borrower is facing worsening conditions like rising wage pressures or raw material pressures or greater competition. We want to know what kinds of surprises do we need to worry about? And how likely are those surprises? Assuming all else equal begs all those critical questions.

There’s a fourth problem with this analysis. The fourth problem is that good management teams and good business owners they adapt to change. We saw this vividly during COVID-19. A sensitivity analysis couldn’t have told you which companies would manage lockdowns well and which ones wouldn’t.

And finally the biggest problem. Even if you get all of the measurements I just went through correct, this analysis it tells you about the impact on your average borrower. Good lenders never lose money on their average borrower, they lose money on their weakest, what statisticians call the tail.

I can tell you right now that Golub Capital’s direct lending portfolio as a whole had a weighted average interest coverage ratio of 2.4 times at December 31, 2022 and GBDC’s weighted average approximate Golub Capital’s direct lending portfolio as a whole. But I don’t honestly think this tells you anything.

What’s a better approach? Well, maybe there are multiple good approaches. But I want to tell you about the approach we’ve used. It’s an approach we’ve tested through multiple business cycles over more than 28 years. And it’s an approach that I think is at the heart of how we’ve excelled and produced premium returns that are consistent over time.

I’ll hand the floor to Greg to explain our approach in detail.

Greg Cashman

Thanks David. I can summarize our approach in one sentence. There is no substitute for granular credit analysis. I mentioned on last quarter’s earnings call that we enhanced our credit monitoring processes in response to the challenging environment.

One of the key benefits of our scale is that when we hit a rough period like the one that we started over the summer, we can deploy some of our 170-plus investment professionals from offense to defense.

We can and did pivot resources from underwriting new deals to scouring the portfolio for potential vulnerabilities. Then we assess each of the companies that we identify as vulnerable one by one.

Specifically, starting last summer, we evaluated company-by-company Golub Capital’s entire middle market loan portfolio. Our analysis focused on six key risk factors.

First, we looked for a portfolio of companies who may have potential liquidity or cash flow issues from increased base rates.

Second, we look for companies susceptible to contracting operating margins. This meant looking closely at cost drivers, as well as pricing power.

Third, we looked at recession resistance. Some companies are more susceptible to recessions than others. We look for companies with material risk of falling revenues, deterioration in working capital, growth in accounts payable and similar vulnerabilities.

Fourth, we look for companies with material vulnerability to a stronger US dollar or to customers or suppliers in areas of geopolitical tension or economic weakness, such as Europe and China.

Fifth, we looked at credits with high levels of EBITDA adjustments that might not be realized in practice.

And finally, we looked at issues specific to our investments in software companies. We look for businesses with material amounts of high-margin, transactional or other non-recurring revenues.

We completed our initial screen around the end of the summer. From there, our work preceded in three different phases.

In phase one, Golub Capital’s Direct Lending Team formed a leadership working group consisting of myself, our Co-Heads of Underwriting and our Head of Workouts. I’ll call this working group DL leadership for short.

DL leadership prioritized an initial set of credits for deeper dives based on the number and magnitude of risk factors that I outlined above. We developed a proprietary portfolio resiliency memo template that probed in detail on the six key risk factors I mentioned. Deal teams completed a resiliency memo for each of the prioritized credits and submitted them to DL Leadership for further review and discussion.

In phase two, deal teams completed and submitted a portfolio resiliency memo for a second wave of credits that were selected by DL Leadership, but which we considered lower priority. In addition, deal teams completed a robust resiliency model for more than 100 credits in order to help us prioritize for phase three.

In phase three deal teams completed and submitted a resiliency memo for a third wave of credits that DL Leadership selected based on the resiliency models from phase two.

In total, 60 portfolio companies underwent a deep-dive bottoms-up analysis across each of phases one, two and three, a further 88 portfolio companies with potential exposure to one or more key risk factors were screened and modeled in phase two and determined not to be priorities for further analysis at this time.

That’s what we did. Now, let’s talk about what we learned. We had two key learnings from our work to-date. First, our analysis led us to conclude that the tail of vulnerable companies in our portfolio is small. Put differently, we saw an encouraging level of resiliency in the substantial majority of portfolio companies we analyzed in detail. Encouraging, but not surprising. Our underwriting process focuses on resiliency and we’re very selective about the companies we lend to.

I mentioned a moment ago that we prioritized 60 portfolio companies for deep dive analysis. Based on the results of our analyses, we determined that 50 of them did not require enhanced monitoring at this time. We did determine that the remaining 10 portfolio companies could benefit from some enhanced monitoring to put ourselves in a better position to identify and address potential vulnerability in the future. As a reminder, these 10 companies were selected from Golub Capital’s entire middle market loan portfolio. For context, seven of these companies are held in GBDC’s investment portfolio and constitute around 3% of the investment portfolio at fair value as of 12/31/22. We believe this illustrates how the size and granularity of GBDC’s portfolio of 300-plus borrowers enables us to mitigate company-specific risks through diversification.

Let me take a moment to explain how we think about the GBDC portfolio companies that we’re monitoring more intensively. In general, these borrowers are currently performing materially in line with expectations, but we concluded we’re at a higher risk of them underperforming prospectively. In general, we expect these borrowers to pay us back in full. Having said this, part of our overall approach to credit monitoring is to be quite proactive to seek to identify vulnerable borrowers early, and to work with sponsors and management teams to increase the margin for error of those borrowers.

You may recall that when COVID-19 hit in March 2020, we went through a similar process of screening the portfolio for vulnerability and prioritizing a subset for enhanced monitoring. Our analysis honed in on the portion of the portfolio in industry sub-segments we believe had relatively significant exposure to COVID-19, such as restaurants, eye care and dental care. We explained that we didn’t necessarily expect this subset of the portfolio to become impaired. In fact, these credits generally recovered in subsequent quarters, and at the same time we thought it was prudent to devote more resources to early detection and if necessary early intervention. We hold a similar view of the GBDC portfolio companies who we’re monitoring more actively as a result of our recent portfolio review.

This brings me to our second key learning, there were common themes among the 50 borrowers that were not designated for enhanced monitoring. Let me give you a few examples of factors that helped us get comfortable with the outlook for these borrowers.

First, we validated that EBITDA add-backs were truly one-off. Second, we saw growth drivers that we expect to improve coverage ratios going forward and that had not been fully reflected in LTM financials, for example, recent acquisitions. Third, we identified credible cost savings and synergies that had not yet been fully realized. Fourth, ample liquidity was available to address temporary operating cash flow shortfalls. And fifth, interest rate hedges were in place, reducing the impact of higher base rates.

We believe these mitigants reflect the care we take in underwriting, especially when we assess EBITDA adjustments. To reiterate a point we made earlier, formulaic stress tests and sensitivity analyses don’t test whether EBITDA adjustments are real in the sense that they’ll roll off and become actual forward earnings power. In fact, we believe the coming period will show that across the middle market some company’s adjustments won’t roll off and won’t prove to be real, and we believe this will be a key driver of increased dispersion and manager performance.

So, to sum up key takeaways from our work to-date. Our detailed granular analysis showed that the portfolio is generally well-positioned for the coming period. We identified a small subset of borrowers that we plan to monitor even more closely than usual. These companies are generally performing okay today, and our expectation is that they will continue to perform. Nothing in these findings leads us to believe that realized credit losses will be outside the bounds of our historical experience.

Two final thoughts. What I’ve described today is part of an ongoing process. We don’t look at the portfolio in detail once and declare mission accomplished. We’ll continue to hone our analysis as more data becomes available. It takes scale and experience to do this well.

And we’re not going to be 100% right. That’s not a realistic goal. Our goal is to detect which of our borrowers are at a higher risk, and then to have early discussions with sponsors and management teams about how to make them more resilient. But there will be surprises. There always are.

We’ll continue to keep you informed about the portfolio throughout this challenging environment. With that, I’ll hand the floor over to Matt.

Matt Benton

Thanks, Craig. I’m going to take the third key question that David mentioned earlier. Will we see more write-downs, or will we see reversals of some write-downs GBDC has already taken?

Let me start by setting context. Over the last calendar year GBDC has performed well, despite a bumpy investment environment. Trailing 12-month ROE was 4.6%, which compares very favorably to traditional fixed income returns over the same time frame. The Bloomberg aggregate index was down 13%, high yield bonds were down 10.6% and the LSTA loan index was down 0.6%.

Our returns were depressed by mark-to-market write-downs in calendar Q2, Q3 and Q4, which were largely a function of credit spreads widening on a market-wide basis and to a lesser extent a function of credit concerns. Net unrealized losses for this nine-month period added up to $0.79 per share, or about 5.1% of NAV as of 3/31/22. Importantly, over the same period GBDC had net realized gains amounting to 0.3% or 30 basis points of 3/31 NAV.

So how should investors think about the $0.79 per share of net unrealized losses? One way to think about it is as an embedded loss reserve.

Let’s look at slide 8 of the earnings presentation. This slide breaks down the $0.79 per share of net unrealized depreciation on investments for the nine-month period ending 12/31/2022 based on our internal performance rating categories. You’ll recall that the highest categories four and five represent loans that are performing as expected or better than expected at underwriting.

The vast majority of our investments fall into categories four and five. They represented 89.3% of the portfolio as of 12/31/22. Critically, the box in gold on the chart shows that investments in categories four and five accounted for 75% of the cumulative unrealized losses incurred since 3/31, or $0.59 per share. If you believe loans in categories four and five are very likely to pay us back as we do, based on our experience, you’d expect GBDC to recapture most of this $0.59 per share over time. Put differently, we think GBDC has a sizable embedded loss reserve attributable to loans that we currently think will be repaid at par.

Second, let’s turn to slide 9. The analysis on the slide essentially looks at GBDC’s actual historical realized loss experience and asks how high would future losses need to be compared to historical experience to eat through what we’re calling GBDC’s embedded loss reserve.

Let me walk you through the chart. The left-hand bar depicts the 5.1% of NAV at $0.79 per share of cumulative net unrealized losses that we took for the nine-month period ending 12/31/2022. The right-hand bar shows GBDC’s worst ever 12-month period of actual net realized losses. This totaled 1.9% of starting NAV for that 12-month period. This means that our unrealized write-downs on the portfolio over the nine-month period ending 12/31/2022 are about three times the size of GBDC’s worst-ever 12-month loss experience.

And bear in mind, GBDC both in the last 12 months and since inception, has recorded average annual net realized gains, not losses. We think the takeaway from this analysis is that GBDC is likely to see reversals of prior write-downs over coming quarters.

With that let’s shift focus to GBDC’s results for the quarter and walk through the earnings presentation in more detail. I’ll ask Chris to kick things off for us and I’ll come back a bit later.

Chris Ericson

Thanks Matt. Turning to slide four, we see GBDC’s adjusted NII per share increased by $0.04 quarter-over-quarter to $0.37 per share, which represents an adjusted NII return on equity of 10%. As David described earlier, the increase in adjusted NII per share was primarily driven by the impact of increasing interest base rates on GBDC’s portfolio and GBDC’s low cost of funding structure. In addition, accrued dividends on certain preferred equity investments generated approximately $0.02 per share during the quarter and will be included in adjusted NII moving forward.

The favorable increase to adjusted NII was partially offset by an accrual for excise tax of approximately $0.01 per share as a result of calendar year 2022 taxable income in excess of distributions or spillover income that was driven by realized gains and short-term temporary book-to-tax income differences that we expect to reverse over time.

GBDC had an adjusted net realized and unrealized loss per share of $0.22, primarily from unrealized depreciation due to a combination of spread widening and isolated credit factors on certain portfolio companies. This unrealized depreciation was partially offset by $0.02 per share of realized gains on the sale of equity investments.

Adjusted EPS was $0.15 per share. We view this as a solid performance in the context of the market and economic volatility and uncertainty.

Now I want to take a moment to quickly provide additional details around two components of adjusted NII. First, we made the decision to incur an excise tax of approximately $2.2 million or $0.01 per share. Historically, we’ve sought to minimize excise tax payments. Our decision this year was primarily driven by the fact that a meaningful portion of the spillover income was related to short-term taxable gains on foreign exchange hedges that we expect to reverse over the balance of calendar year 2023. We felt this was more beneficial than NAV stability over calendar 2023 versus paying out a large one-time special distribution on a meaningful portion of spillover income that will ultimately reverse.

Second, we evaluated our practice for recognizing dividends on preferred equity investments which previously were recognized in unrealized appreciation. This component of the portfolio reached a large enough size of our calendar year 2022 that we felt it was appropriate to accrue these dividends within net investment income, which is consistent with industry practices. This added $0.02 per share to adjusted NII and will be a continuing component of adjusted NII moving forward.

One other comment that I would make here – the majority of these preferred dividends are structured as non-cash or PIK income and are generally collected upon redemption and equates to approximately 2.5% of our overall investment income.

Turning to slide 7 you can see that NAV declined 1.2% quarter-over-quarter to $14.71 per share from $14.89 per share. Let’s walk through the components. As I just mentioned, adjusted NII was $0.37 per share, and the company paid $0.33 per share of dividends. Adjusted NII was offset by a loss of $0.23 per share from the net change in unrealized depreciation on investments. And finally, net realized appreciation came to a gain of $0.02 per share.

We turn to slide 12. This slide summarizes our origination activity for the quarter. Net funds growth increased slightly quarter-over-quarter, primarily due to new investment commitments, and delayed draw term loan fundings exceeding exits and sales of investments and the net change in fair value of investments. The asset mix shown in the middle of the slide remained fairly consistent with our prior quarter originations.

Looking at the bottom of the slide, the weighted average rate on investments increased by 210 basis points this quarter from a combination of higher base rates and wider assets spreads on new originations. The weighted average spreads on new investments increased by 50 basis points over the prior quarter from 6.2% to 6.7%.

Slide 13 shows GBDC’s overall portfolio mix. As you can see the portfolio breakdown by investment type remained consistent quarter-over-quarter, with one stop loans continuing to represent around 85% of the portfolio at fair value.

Slide 14 shows that GBDC’s portfolio remain highly diversified by obligor, with an average investment size of approximately 30 basis points. As of December 31, 2022, 94% of our investment portfolio was comprised of first lien, senior secured floating rate loans and defensively positioned in what we believe to be resilient industries.

Turning to slide 15, as we explored in detail last quarter, the rising interest rate environment really highlights the asset sensitive nature of GBDC’s balance sheet. Let’s start with the dark blue line, which is our investment income yield, which includes the amortization of fees and discounts. GBDC’s investment income yield increased by 130 basis points, primarily from rising interest rates. By contrast our cost of debt, the teal line, only increased 70 basis points. Our cost of debt benefits meaningfully from our $1.5 billion of unsecured notes that are fixed rate and have a weighted average coupon of 2.7%. Combining these two factors, our weighted average net investment spread, the gold line, increased by 60 basis points over the prior quarter.

I’ll now turn it back over to Matt to discuss how GBDC is positioned for higher rates. Matt?

Matt Benton

Thanks, Chris. On slide 16, we’ve quantified the potential positive impact of higher base rates on GBDC’s NII earnings power. The key takeaway is that GBDC’s adjusted NII per share stands to continue to benefit from two key tailwinds in the coming period.

The first tailwind is that there’s a lag between when base rates change in the market and when loans in our portfolio reset the higher base rates, which happens once per quarter in those cases. Said another way, GBDC sees the full benefit of higher base rates about a quarter after those base rates actually increase. The chart on this slide is our attempt to help demystify this dynamic for you. As you can see on the chart, the average LIBOR or SOFR rate GBDC actually earned on its investments for the quarter ended 12/31 was meaningfully less than market rates at the end of the quarter.

The second tailwind is that base rates will increase further from 12/31 levels based on the recent Fed decisions and the expectations embedded in the forward curve.

The leftmost bar shows GBDC’s actual adjusted NII of $0.37 per share for the quarter ended 12/31. On average for the quarter, the average LIBOR rate was approximately 375 basis points. The right bar looks at what GBDC’s adjusted NII per share would have been in the 12/31 quarter if all of its floating-rate assets and liabilities had been based on a LIBOR rate of 477 basis points, the rate at quarter end. In this scenario, all else equal, we estimate adjusted NII would have increased 13% to $0.42 per share. Please note that today’s three-month LIBOR is a bit above the level we assumed in this analysis.

The bottom line is that we think GBDC’s NII per share has a lot of built-in momentum just from higher base rates that have already occurred but not yet flowed through to GBDC’s results. We aren’t assuming wider spreads on existing investments for example, from amendment activity or increased payoffs. In our view those are likely drivers of further NII upside. You’ll see additional details on GBDC’s asset sensitivity in the Form 10-Q if you’d like to drill down further.

Let’s move on to Slide 17 and 18 and take a closer look at credit quality metrics. On Slide 17, you can see the number of non-accrual investments as of 12/31 increased to 9 from 8 compared to 9/30. This is because the disposition of one non-accrual investment was offset by the addition of two new non-accrual investments. Additionally, the percentage of investments on non-accrual measured at fair value increased modestly from the 9/30/22 quarter to 1.8% of our total portfolio from 1.3%.

On Slide 18, as David mentioned earlier, internal performance ratings have been strong and stable and consistent with pre-COVID-19 levels. Over 89% of investments have an internal performance rating of four or higher, which means they are performing as expected or better than expected on underwriting and only 1.3% of investments have an internal performance rating of two or lower, which means they are performing materially below expectations at underwriting.

We’re going to skip past Slide 19 through 23. These slides have more detail on GBDC’s financial statements, dividend history and other key metrics. The last slide I want to cover before handing it back to David is Slide 24. We believe GBDC has meaningful embedded value in its funding structure. We ended the quarter with almost $750 million of dry powder from unrestricted cash, undrawn commitments on our meaningfully over-collateralized corporate revolver and the unused unsecured revolver provided by our Advisor.

Our GAAP debt-to-equity ratio as of 12/31, net of unrestricted cash, was 1.19 times. 47% of our debt funding is in the form of unsecured notes, the majority of which have maturities in 2026 and 2027. We issued these fixed rate notes with a weighted average coupon of 2.7% and did not swap these out to floating rate.

Our weighted average cost of debt for the quarter ended December 31, 2022 was 4.4%, which we believe is among the lowest in our peer group of publicly traded BDCs. We believe our funding structure is a meaningful competitive advantage.

I’ll turn it back over to David for closing remarks and Q&A.

David Golub

Thanks, Matt. To sum up, GBDC’s performance for the quarter ended December 31 was solid. Adjusted NII per share was strong and it was well in excess of our recently raised dividend. The portfolio is generally healthy and it’s performing well from a credit perspective. Unrealized losses have been elevated for the last several quarters, but as we’ve said many times before, what really matters in the long run is avoiding realized losses. And in the most recent quarter and in the last nine months, we once again reported net realized gains.

We’ve always believed that early detection of risks and early intervention to mitigate those risks are critical for limiting credit losses. It takes scale and experience to do this well. We’ve today taken you through how we last year undertook what we believe was a very thorough review of the portfolio against a range of key risk factors and how we honed in on a small subset of names that we plan to continue to monitor quite closely. Now Greg said it right. We’re not going to be 100% right. That’s not a realistic goal. Consistent with prior periods, our approach is to try to identify borrowers that are higher risk and then to have early discussions with sponsors and management teams about how to make them more resilient. This approach has worked before and I believe it’s going to work again. What do I mean by work? I mean that once again we’ll be able to keep realized credit losses low and we’ll see a large portion of the unrealized losses that we’ve taken over the last nine months reverse.

Finally, I promised I’d come back to the topic of the dividend in my closing remarks. We out-earned our dividend by $0.04 per share in fiscal Q1. And as we’ve said before, we think GBDC’s earnings are going to be driven higher by higher base rates, higher spreads and GBDC’s very low cost of funds. We’re evaluating in this context whether and when it would be appropriate to increase the quarterly dividend further or make a supplemental dividend or both. For now, we think it’s prudent to wait and see, but we’ll keep you informed as our thinking progresses.

With that, let’s open the line for questions.

Question-and-Answer Session

Operator

Thank you. [Operator Instructions] We’ll take our first question from Finian O’Shea with Wells Fargo Securities. Your line is now open.

Jordan Wathen

Hi. This is Jordan on for Fin today. Just a question on your loan documents. You spotted some PIK on – you spotted new or higher PIK on about 3% of loans this quarter. Is this — if you could just like characterize this? Is this something that’s normal course toggling? Maybe default interest or something else entirely?

David Golub

Combination of all of the above, Jordan. So we are seeing, particularly in our Golub Growth portfolio, demand from sponsors for loans that have a PIK component or an optional PIK component. So that’s part of what you’re seeing. And there are a couple of cases where we have added a PIK component to existing loans as part of amendments, as part of efforts to bring loans that were under-priced up to levels that we were looking to get them at without creating more cash drag on the companies.

Jordan Wathen

Okay. That’s helpful. And then question for Chris. Did the BDC pay a full incentive fee back to the advisor this quarter? And if so what’s kind of the — what is the cushion before earnings fall back into the band?

Chris Ericson

Yes, hi. We did pay a full incentive fee this quarter and we’re well-above the 8% hurdle rates. I’ll have to get back to you with the exact number on that.

Jordan Wathen

Okay. Thank you so much. That’s it from me.

Operator

Okay. Next we’ll go to Robert Dodd with Raymond James. Your line is now open.

Robert Dodd

You gave about the portfolio vulnerability analysis, et cetera…

David Golub

Robert, can you just start again? Yeah. Can you just start your question again?

Robert Dodd

Yeah, sorry. It’s tied to the vulnerability analysis that you completed. The disclosure is very helpful, the information in the presentation. How much of that if any was, is normal per share with say your third-party valuation consultant each quarter into the fair value analysis that they assist with? Or was this all incremental over and above the kind of normal analysis that goes into evaluating what the appropriate fair values are for the loans each quarter?

David Golub

So we do, just to set context for everyone on the line, each quarter Golub Capital has an internal group that does valuation work with respect to every position in the portfolio and approximately 30% of those valuations are also reviewed each quarter by third-party valuation experts like Duff & Phelps, Houlihan Lokey, or Kroll, Houlihan Lokey, et cetera.

The information that we share with the valuation firms is a robust set of information. So there’s nothing that we’re looking to prepare, Robert, that we would not be sharing with the valuation firms. But the analysis, the resiliency analysis that we went through is a special analysis in the sense that we began in the late summer of last year to redeploy a significant portion of our underwriters to do an exercise that we’ve done before. We did it during COVID, we’ve done it during other periods of rapid change, where we go through and screen the portfolio again to look at vulnerability to changes that we’ve identified.

So I think the answer to your question is, it is both consistent with information that’s provided to the valuation firms and it’s a new analysis that we do episodically as opposed to every quarter.

Robert Dodd

Got it. I appreciate that. And then if I can one more, on kind of on credit, setting the landscape again. I mean, your non-accrual rate is not high by industry historic standards. So I just want to say that. Historically it’s been well below industry historic standards as well. But it is now, if I look at your non-accrual at cost, for example, or you know, fair value, it’s relatively elevated by your standards historically. About the only level higher than this in the last more than 10 years was the peak during COVID. So can you give us any more color about what’s — what — it’s not many assets and the rate is objectively not that high, but why is it so high compared to your normal low levels currently?

David Golub

So you’re focused on the percentage at cost. I don’t focus on percentage at cost. I focused on percentage at fair value. I think that’s the amount that we have at risk at a given point in time vis-à-vis the difference we’ve already taken that as a hit against earnings. And if we look at the non-accrual investments as a percentage of fair value, it’s not out of line with our history. It’s not — it’s low by the standards of our industry.

We do have a couple of credits in this pool where our current carrying value is small. And I think the cost of those investments is elevated relative to fair value and that’s kind of creating some distortions when we look at the percentage vis-à-vis cost. We’re going to go back and do some more work on this and come back with some more analysis that I hope to be able to share with you.

Robert Dodd

Okay. Appreciate it. Thank you.

Operator

And next we’ll go to Ryan Lynch with KBW. Your line is now open.

Ryan Lynch

Hey. Good afternoon. First question I had — well, let me just first say, I really do appreciate all the detail you guys put in and walk through kind of the resilient analysis and what you guys discovered as well as the couple of slides that went over kind of where the markdowns have come in your portfolio versus kind of potentially weaker credits versus kind of maybe mark-to-market declines that could recover. So it’s a really great analysis, really appreciate that.

On the resilient analysis when you talked about the seven investments that you guys identified that consist of in GBDC’s portfolio. I’m just curious I think you said that generally those investments are performing fine today, but there’s additional heightened potential for weakness down the road.

Have you already started having conversations with the sponsors of those positions yet? And also, from just a higher level, you mentioned having this robust detection process helps early detection and early mitigants to risk. Can you just walk through, when you guys have identified a risky investment and it does start to struggle, what are the things that early detection and risk mitigant, what are some of the strings you can pull to help increase the recovery potential of an investment that’s struggling?

David Golub

Sure. Thanks, Ryan, and I appreciate your positive feedback on the new approach to our earnings presentations and the new information. Thank you. Look one of the characteristics of Golub Capital that’s unusual is we are almost always the lead lender in our loans.

We — I think it’s over 90% and it’s been over 90% for many years in a row. In a typical deal we’re, if not the sole lender, we’re in a small bank group that we’re the lead lender of. So that puts us in the position to be able to do things that are very challenging to do in the broadly syndicated market or even in the private market in situations where you’re a participant in a larger club.

It puts us in a position to have conversations with the borrower and with the sponsor where they know and we know that we’re the decision-maker. And they know and we know that there’s an opportunity to be creative and thoughtful about solutions. And in most cases this is with a sponsor we’ve done multiple deals with. So they know and we know, we’re both interested in sustaining a strong and positive relationship on an ongoing basis so that we can continue to do business together.

So with that by way of context, we’re in constant touch with our borrowers and our sponsors. It’s not just on poorly performing companies; it’s on all of our companies. When we have a company that we’re growing concerned about, we have more discussions with the borrower and with the sponsor. And we’ll talk about concerns that we have. We’ll talk about their plans and ways in which they can potentially take steps that might increase a margin for error. That might mean slowing down an expansion plan or decreasing CapEx or selling a division or a couple of small pieces of the business, or it might involve the sponsor putting in more capital, or in some cases it may involve putting the business up for sale.

These are all discussions that we regularly have with sponsors and with management teams. And those discussions are informed by the credit monitoring work that we do.

I think this goes in sharp contrast to say the broadly syndicated market where you have a very large collection of lenders, no one of which is a decision maker, and where the borrower doesn’t really even have a reasonable expectation of being able to get that group to agree to make meaningful changes because they’re so frequently a group of lenders or a single lender that won’t agree. So it’s a very different dynamic Ryan. It’s purposefully a setup where we have more power and more control. And I think it works to the benefit of everybody involved.

Ryan Lynch

Okay, that’s helpful background and color on kind of a little bit of insight into those processes and conversations. The other question I had and it kind of came up when you were talking about the weighted average interest coverage and how valuable that is.

You all publish the Golub Capital Altman Index and I think it’s very helpful to kind of give an insight on how general trends in the portfolio are moving on a quarterly basis. But when I think about your comments on kind of the weighted average interest covered not being a good metric because the average borrower is not going to really struggle, it’s going to be kind of those marginal credits. Could you take that same sort of approach when you look at the Golub Capital Altman Index and say maybe the results that you look at that on a quarterly basis aren’t as helpful because it’s just showing the average and it really may not be indicative of the true credit performance of those individual underlying borrowers in your portfolio?

I’d just love to hear your comments on this. I find the index really helpful. But in context when you said about the weighted average EBITDA, I’m just wondering — I’d be curious to hear your thoughts on that.

David Golub

So it depends what you’re using it for. I think your point is well taken. If you’re thinking about using the index to create an indicator as a metric of how the portfolio is doing as a whole, I think it’s very useful.

If you’re looking at the index as an indicator for how the tail is doing, I don’t think it’s very useful. I think the best source of information about the tail is our portfolio ratings. And one of the things that I like to emphasize in discussions with investors is are we seeing significant migration toward categories one and two, those are the two risk performance rating categories where — which contain investments that are significantly underperforming. Are we seeing more and more of the portfolio in those two categories?

What you would expect in a pre-recessionary or recessionary environment is that kind of credit migration. You’d see quarter-over-quarter increases repeatedly over a series of quarters and you’d expect to see levels of category one and two credits that are high by historical standards.

And I think — what we can say is on a portfolio-wide basis, the middle market index numbers are encouraging. And in respect of the tail, the portfolio performance ratings are encouraging in that they do not show the kind of migration that I was alluding to.

Ryan Lynch

Okay. That’s helpful and I think makes a lot of sense. That’s all for me today I appreciate the time.

Operator

[Operator Instructions] I’m showing no further questions. I’ll turn the call back over to David Golub for any additional or closing remarks.

David Golub

Thank you. I just want to thank everybody for their time today. I know this was a particularly long presentation so thanks for your patience. We did want to present to you all some very detailed information about the portfolio and our approach to credit monitoring. So I appreciate your patience on that.

As always also appreciate your partnership. Should you have any questions, always feel free to reach out and look forward to talking again next quarter.

Operator

This concludes today’s conference call. You may now disconnect.