Description

Summary Thesis

Ares Capital Corporation is a Business Development Corporation (BDC) with a strong track record that extends through the last cycle. It is in the midst of several major transitions that are causing an overhang on its share price. As it moves through these transitions over the next year, the stock should rerate 10-20% while shareholders collect a 10% dividend.

ARCC has been written up on VIC twice before (and there have been 20+ BDC write-ups), so this write-up assumes a basic familiarity and focuses on: (1) our framework for evaluating BDCs as credit investments; and (2) the circumstances surrounding ARCC that make this idea timely.

Evaluating BDCs as High-Yield Alternatives

The yield premium in middle market direct lending (where BDCs like ARCC play) is large and real. By analogy to the broadly syndicated/corporate bond markets, these loans can offer B/CCC yields on BB credits—a premium of 5-6%. There are a number of reasons underlying this premium:

• Size and illiquidity are the most significant factors. These factors are well-recognized in the equity markets, and there have been studies showing that they also matter in the credit markets (see, e.g., Elton, et. al., The Journal of Finance, 2001, p. 247).
• Origination fees add to the excess returns of direct lending—the fees that would go to an investment bank in a corporate bond issuance accrue to the lender instead.  Depending upon the deal, this can add 2-4pp additional return, which gets amortized over the ~4 year life of the loan.
• Additional fees (amendment, prepayment, syndication, etc.) are irregular but can be meaningful.
• Tight covenants and the direct, simplified lender-borrower relationship allow the lender to exert greater control in stressed scenarios. This can reduce the cost and timeframe of restructuring while lowering loss severity in the event of a default.
• Finally, active management can be a source of excess return—this can be through a manager’s investment skill, but there are also some “simple” ways in which active management can add value (e.g., turning over investments above par).

The yield premium in direct lending is validated by fact that this asset class is institutionally accepted and rapidly growing. Institutional direct lending is experiencing long-term tailwinds as regulation (Dodd-Frank, CLO risk retention, etc.) reduces some of the traditional sources of credit supply.

However, not all managers are able to capture this large yield premium, and BDC credit performance can be difficult to assess, as capital gains and different payout structures can mask credit losses. The worst managers are undifferentiated from the CCC market, with credit loss rates as high as 500bps. ARCC has one of the strongest track records, with a gross loss rate of 150bps per annum over its GFC-spanning 11-year existence (though the apparent track record is even better, as these losses were fully offset by capital gains).

ARCC’s returns (IPO to date; dividends plus NAV accretion) have been over 12% per annum. Roughly 1% p.a. of this return can be attributed to accretive share issuance, while another 1% of this return came from its 2009 bargain purchase gain of Allied Capital (ALD). Another 1% came from the gradual work-out of impaired positions acquired from ALD, such that the “fundamental” historical ROE of ARCC has been ~9.5%.

BDC fees are well above market for institutional direct lending but are still less than the excess returns generated. Institutional fee structures for middle market direct lending are typically 1/10 or 1/15, with administrative expenses of 10-25bps. This works out to 20-25% of the 10-11% gross returns or 30-40% of the loss-adjusted excess returns over a ~4% loss-adjusted high-yield benchmark.

In contrast, most BDC fee structures range from 1/20 to 2/20, with administrative expenses of 50-150 bps on gross assets. At the extreme, BDC fees and expenses eat up all the excess return—why pay par for such a product? ARCC’s fee structure is more moderate; it works out to 3.5% of gross assets in the average year. This is over 30% of the ~11% gross returns that ARCC is generating and 60% of the excess gross returns over a ~4% loss-adjusted high-yield benchmark. The fees are high, but there is still ~2pp of excess performance over high yield (50% over the benchmark!).

The permanent-life structure of BDCs is value-additive. In contrast to defined-life direct lending funds, there are no start-up or wind-down costs. More importantly, a well-managed BDC can accretively raise equity when the shares are trading above book and repurchase equity when it is trading below NAV. ARCC has created an additional 1pp of value per annum through accretive issuances.

Even though BDCs are “yieldy” assets, rising rates are generally a positive for them. The average maturity of a loan in a BDC portfolio is around three years, and most of those assets are floating-rate. BDC earnings will rise in a rising rate environment.

To a first approximation, the valuation case is simple—ARCC holds a dynamic pool of loans yielding 11-12%. You pay 3.5% to access this proprietary pool, giving a net yield of ~8% (on assets). High-yield with a similar loss rate has a 5-6% net yield. Absent idiosyncratic issues, the NPV of this yield premium suggests that fair value for ARCC should be 20-30% over par.

Idiosyncratic Issues at ARCC

Given ARCC’s size, some of the highest-alpha opportunities at the smaller end of the middle market are no longer available/ material to them. However, this has been an issue for some time, and they have been thoughtful about adjusting their model to mitigate this issue—using their scale to their advantage wherever possible. In particular, they have tweaked their model to be more fee-dependent than other BDCs—through syndication fees, structuring fees, and through Ivy Hill Asset Management. This strategy allows ARCC to play to its strengths, and these high-quality, fee-related earnings carry no credit risk and diversify the income stream.

Ivy Hill is an asset manager owned directly by ARCC. It accounts for 3% of the portfolio but 4% of TII. Ivy Hill manages low-levered, direct-lending CLOs for third party investors. Most of its income comes from management fees, though it does hold some CLO tranches on its own balance sheet (enough to be risk-retention compliant).

The Senior Secured Lending Program (SSLP) is a structure pioneered by ARCC to increase its effective leverage. By law, BDCs are required to fund with no more than 1:1 debt-to-equity; however, they are able to invest up to 30% of their balance sheet in “non-eligible assets,” including leveraged pools of debt. To this end, there are some BDCs that own CLO equities. ARCC’s SSLP is similar to the junior tranche of a CLO, but it is structured as a JV with GE finance in which GE holds the senior tranche (along with a small interest in the junior tranche alongside ARCC). At 3:1 to 4:1 leverage with the senior tranche priced at ~5%, the SSLP enabled ARCC to transform ~7% loans into a security yielding ~14%. This structure has been so successful that a number of other BDCs have copied it.

Unfortunately, GE wants out of this JV as part of its mandate to exit the finance business. As SSLP loans are repaid, the cash is being returned, first to the senior tranche, then to ARCC’s junior tranche. This means that >20% of ARCC’s balance sheet is invested in a security whose yield has fallen from ~14% to ~8.5%, which has been a drag on earnings. The SDLP is a nascent JV with AIG meant to replace the SSLP. The structure is similar, and they launched the SDLP in Q3 2016. This complicated scenario has created several issues that concern the market:

• Execution risk—there has been some concern that ARCC would fail to get the SDLP off the ground. This risk has been addressed by the successful launch last quarter. There is a related concern that the SDLP’s growth will be stunted by the “30% rule” that mandates that no more than 30% of their balance sheet can be invested in “non-eligible assets,” such as the SSLP and SDLP. We view this concern as a non-issue, as the SSLP should be in a position to return cash soon, and there is no consequence to postponing the growth of the SDLP several quarters, aside from…
• Depressed earnings—the low SSLP coupon and the warehousing of lower-coupon loans destined for the SDLP has depressed ARCC earnings over the past year. However, the launch of the SDLP last quarter marked the low point, and—even at this low point—ARCC’s 10% dividend remained covered.
• Increased leverage—our main concern with the SSLP & SDLP is how they increase ARCC’s effective leverage: 3:1 leverage in an SDLP that accounts for 25% of the balance sheet raises the BDC’s effective leverage from 0.7:1 to 2:1. This is higher that we would like, and we favor BDCs that don’t have this shadow leverage. However, there are two mitigants that allow us to get comfortable with the increased volatility that this structure brings: (1) the loans in the SSLP & SDLP are lower coupon and lower risk than the average BDC loan; and (2) we are paying much lower fees on the SSLP & SDLP assets. For example, if the SDLP is levered 2.5:1 and BDC expenses run 3.5% of gross assets, we are only paying 1% on the look-through assets of the SDLP, meaning the 7% SDLP loan yields 6% to us—the equivalent of a 6% + 3.5% = 9.5%-yielding loan held at the BDC level.

A deep dive into a BDC’s portfolio will always reveal several “concerning” positions, and ARCC is no different.  On the top of ARCC’s “concerning positions” list is its preferred equity stake in Iniflaw Holdings, an operator of for-profit law schools. They finally started marking this down in Q3, though it still accounts for $90mn on the balance sheet. They also have an impaired $60mn position in a major for-profit educator in Puerto Rico. These are both on nonaccrual and account for ~2% of ARCC’s NAV. We are also concerned about their position in CCS Intermediate Holdings, which provides healthcare and dental services to prisons and accounts for ~2% of ARCC’s NAV.  It is worth noting that for-profit education and for-profit prisons are some of the Obama administration’s most-despised industries, and the regulatory environment for these businesses should improve under the next administration. It is also worth noting that ARCC’s $125mn senior secured loan to Primeex Energy had been near the top of our “concerning positions” list until last quarter, when it was paid off in full (it had been carried at 90 the previous quarter). Overall, we view ARCC’s NAV as overstated by ~3%.

ARCC’s pending acquisition of American Capital (ACAS) is viewed with some amount of market suspicion, as the interests of management (grow AUM) and shareholders (grow NAV & dividend) are not aligned. This view is supported by management’s projections of NAV dilution, combined with their low level of disclosure on the pro forma entity. We disagree with this view:

• Based on our diligence of the ACAS portfolio, we think there is upside to the marks. ACAS realizations have tended to be positive for many years. Positions that were specifically highlighted by ACAS short-sellers as suspect (e.g., SEHAC) were sold above mark. In addition, ACAS took significant marks during its strategic process, and ARCC marked ACAS down a further 5-10% in the proxy. A moderation of the proxy markdown would make this deal NAV-accretive. Furthermore, we note that the gradual markup/realization of these legacy ACAS positions should provide a nice earnings tailwind for ARCC shareholders in the coming years (not unlike the gradual markup of legacy ALD positions).
• As noted above, we also believe that ARCC’s portfolio was over-marked in the proxy. Applying these adjustments to the merger math makes the transaction NAV-accretive.
• ACAS pays no dividend, and it is therefore accruing value to ARCC until the merger closes. Nine months (3/31/16-12/31/16) of 5% ROE on $4.0bn equity is $150mn—enough to make the transaction NAV-accretive by the close.  ACAS has already generated $123mn NOI in Q2 & Q3, and their reported net increase in NAV over those two quarters is double that figure.
• The additional scale from the deal should improve ARCC’s G&A leverage by 5-10 bps on gross assets. While this is small, it should structurally increase ARCC’s earnings power by 1%.
• By expanding ARCC’s balance sheet, the merger expands the size of ARCC’s “30% rule bucket,” allowing it to grow the SDLP more rapidly while the SSLP winds down. While this is only relevant in the short term, it does enable ARCC to replace 10-11% paper with ~14% paper, accelerating its earnings growth.
• In summary, this deal is NAV-accretive by several percentage points and even more accretive to earnings.

We believe poor investor communications have exacerbated some of these overhangs on the stock. Like many BDCs, they are loathe to discuss credit issues. They will tell you that they aren’t exposed to CLOs, even though Ivy Hill is a CLO manager. They will point you to their net loss record (credit losses + equity capital gains) and obfuscate their gross credit loss record (even through that track record is actually very good). They insist upon calling GE’s senior securities within the SSLP “partnership interests” rather than senior debt. With respect to the pending ACAS merger, they have been slow to share even the most basic details (such as NAV accretion math). This has fed investor concerns over misaligned management.

Valuation & Conclusion

We think ARCC’s true NAV lies a bit over $16.00 and will rise to $16.50 after the ACAS merger closes in January. The 30% premium-to-par price target suggested by the 2-3% post-fee yield premium seems aggressive (we are concerned about ARCC’s exclusion from the lower middle market and how late we are in the cycle), but a premium-to-par of half that seems appropriate—implying an $18.50 price target. The dividend at this price would still be above 8% (and a 9%+ yield based upon earnings power). This price target is 20%+ up—suggesting a 30% one-year return—very attractive for a highly liquid credit investment.

Addendum: we spend a fair amount of time on BDCs—happy to discuss the sector generally and answer any questions. To preempt the two most obvious questions: (1) we view the sector as slightly undervalued at current prices, but nothing to get excited about (in contrast to early in the year); and (2) other names that we think are attractive here include: CSWC, TPVG, TSLX, and TCRD.

Catalyst

• Close of ACAS merger
• Ramp of SDLP & cash return from SSLP
• Earnings and dividend growth