After a tumultuous last couple of quarters, Oxford Lane Capital (NASDAQ:OXLC) should return to a more stable footing during the second quarter of 2016. With the sale of investments to repurchase the preferred shares and the continued improvement in CLO equity valuations, OXLC's asset coverage ratio test should be satisfied as of June 30, 2016. Based upon my calculations, the asset coverage ratio should be roughly 2.1x (assuming no additional repurchases of preferred shares), which will allow management to declare a third-quarter dividend. In addition, I expect NII of $0.36 and Core NII of roughly $1.15 for the second quarter, so I assume that the third-quarter dividend will be the standard $0.60.
In the article, I reference figures or other information contained in OXLC's first-quarter investor presentation (1Q report). It would be helpful for the reader to have access to the first-quarter report while reading this article; special attention should be paid to pages six and nine. In addition, it may be helpful for the readers to reference my in-depth CLO equity accounting and valuation article, "Risks and Rewards of Investing in CLO Equity Funds: A Case Study Analyzing 4 BDCs and CEFs."
OXLC's historical NAV
In this article, I will examine the following items:
Sale of investments - Analyze the details of the first-quarter sale of CLO equity investments and the impact of those sales will have on OXLC's future performance; second-quarter projected results - I will provide projections for the second-quarter results; and CLO performance - I will examine the performance of the overall CLO market and I will examine certain key CLO performance matrices provided by OXLC and compare them to industry averages (as provided by Wells Fargo's research report dated July 22, 2016, "The CLO Manager Style Guide") ("WF CLO Report"). I will also attempt to show the reader how these matrices can provide them with valuable insight into OXLC's current and future performance, especially the amount and sustainability of future cash distributions.
Sale of Investments
Since the management discussed the sale of investments in the first-quarter conference call, I don't plan on spending a significant amount of time analyzing the sale. However, I do believe there is some important information that can be gleaned from the details of the sale. Management sold three large positions and several smaller investments. I am sure management considered a plethora of variables in deciding which investments to sell - liquidity, price, underlying performance of the CLO equity, relationship with the CLO manager, etc. While it is difficult to ascertain the underlying credit quality of these investments, it appears that management took advantage of the opportunity to sell poorly performing deals, at least as measured by the drop in its fourth-quarter fair value. Seneca Park CLO's, OZLM VII's and Telos CLO 2013-3's fair values dropped significantly more than the average CLO Equity investment during the fourth quarter; 20.1%, 17.1% and 16.0%, respectively, compared to the average fair value drop of 11.4% (see fourth-quarter 2015 OXLC investor presentation, page 9).
OXLC sold the investment for approximately $34.9 million, with an amortized cost of $55.5 million, causing a realized loss of approximately $20.6 million, or $1.13 a share. Based upon my calculations, the fair value of the investments sold were roughly $43.6 million at December 31, 2015. The sale price of $34.9 million represents a 20% drop in the fair value during the first part of 2016.
First-quarter sales and purchase
Investments sold had a Weighted Average Effective Yield ("WAEF") of 15.3%, which is very close to the overall WAEF on the CLO equity portfolio of 15.4% at December 31, 2015 (see page 8 of OXLC's first-quarter investor presentation). The loss of $55.5 million of CLO equity investments (measured at amortized cost) will have a significant impact on OXLC's interest income. The sale of these investments will reduce cash distributions by approximately $4.0 million and interest income by approximately $2.1 million a quarter. These losses will be partially offset by the reduction in the interest expense on the repurchased preferred shares of approximately $0.5 million per quarter.
In my last article on OXLC, I think I was crystal clear about my dim view of management's decision to sell investments at the bottom of the market (or at least very close to it). After reviewing the details of the sale, my overall impression is unchanged, but there are some positives to take away from the sale. First, based upon my back-of-the-envelope calculations, I don't believe Oxford Lane's asset coverage ratio would have been satisfied at June 30, 2016 without the repurchase of the preferred shares. Therefore, the sale did provide management with the opportunity to declare a third-quarter dividend. Second, it should be noted that management sold approximately $10 million of investments in excess of what it used to repurchase the preferred shares. I am confident that OXLC was able to use this $10 million (plus the additional $10 million of cash on its balance sheet at March 31) to either (X) invest in attractive CLO equity investments, which should be very accretive to earnings over the long term, or (Y) repurchase additional preferred shares to provide it with more breathing room with its asset coverage ratio.
Second-Quarter Projected Results
During the second quarter, the CLO equity market seems to have improved quite a bit. With the sale of investments behind the company and the improvement in CLO equity valuations, OXLC should also be entering a period of stability where its financial results will become more consistent and less volatile. Based upon Eagle Point Credit Company (NYSE:ECC) published preliminary NAV at June 30, 2016, I am projecting a 10% increase in the valuation of OXLC's CLO equity portfolio. Given this, I believe Oxford Lane's asset coverage test should be satisfied at June 30, 2016; therefore, management will be free to declare a third-quarter dividend. In calculating my second-quarter results, I assumed that management invested the $20 million in new investment, yielding roughly 20% and that it did not repurchase any additional preferred shares. If management did repurchase preferred share, then obviously its NII will be lower, but its asset coverage test will be higher than my projected 2.1x.
I am projecting NII of approximately $0.36 a share (range of roughly $0.32-0.38) and Core NII of $1.15. The NII is based upon a WAEY of 14.3%. The relatively strong Core NII is due to the fact that Atrium XII ($42.8 million par) and OZLM XIV ($13.5 million par) had their first payment dates during the second quarter. The additional cash generated from these two investments should be able to offset the loss of cash distributions on the assets sold (at least for the second quarter).
Both Atrium XII and OZLM XIV had extremely strong initial equity distributions, both CLOs distributed roughly 17.5% based upon their par balance. The large distributions represent both the normal net interest distributions and trading gains earned during the loan accumulation phase and during the initial phase of the CLO. Loan accumulation funds are privately negotiated deals with no standardized rules, so I can only guess that trading gains earned during the loan accumulation period where included in the distributions. OXLC invested in both of its loan accumulation funds. Also, the two CLO managers, Credit Swiss Asset Management (CSAM) and Och-Ziff Loan Management, are both known as extremely savvy loan investors and have a strong record of returning outsized returns to the CLO equity investors.
Based upon the new CLOs coming online this quarter, I projected cash distributions on CLO equity of $28.0 million or 6.6% of par. OXLC's CLO equity distributions should return a normalized distribution of roughly 5.0-5.5% going forward.
Just a quick note on future CLO equity distributions, I happened to come upon the amount of AIMCO CLO 2014-A (OXLC owns $26 million par value of the equity) equity distribution that occurred in the third quarter and it was 4.13% (17.26% divided by 4), which compares favorably to its first-quarter distribution of 3.8% (see page 9 of OXLC's first-quarter investor presentation).
Effective Interest Rate
OXLC's WAEY has stayed somewhat stable over the last three quarters; 14.3%, 15.4% and 13.9%, respectively. However, the effective yield for each CLO equity has shown much more volatility (see chart 4 below). As I have discussed in several of my CLO-related articles, there appears to be two distinctive accounting methodologies used for calculating effective yield. The first method is to constantly monitor the performance of each CLO and adjust the assumptions that go into the effective yield calculations. The second method is that management initially takes a long-term view of the assumptions that are used to calculate the effective yield (due to the long-term nature of the assumptions, they are generally more conservative) and only update those assumptions when it is apparent that they are no longer valid. OXLC uses the first method while ECC uses the second method. As I have stated before, I am completely agnostic to which methodology that a fund employs, however, an investor should be aware of the differences in the accounting methodology. Given its accounting methodology, OXLC will naturally have more fluctuations in its calculated effective yield than ECC. One upside of OXLC's methodology is that the changes in a CLO equity's effective yield will give the investor insight into management's thoughts on the performance of each investment.
Change in Effective Yield
The investor can assume that OXLC's management believes that Carlyle Global Market Strategies CLO 2013-2 is performing better than expected (an increase in effective yield of almost 9%) while B&M CLO 2014-1 is performing below expectations (a decrease in effective yield of 10%).
Depending on how OXLC's management invested the available $20 million in cash, I believe that it will earn approximately $0.36 of NII and Core NII of $1.15 for the quarter. Absent any severe downturn in the leverage loan credit cycle, OXLC should be able to sustain its $0.60 quarterly dividend for the next few quarters at the least. In the following section, I will analyze the CLO performance metrics that OXLC provides to its investors and how an investor can use this information to gain greater insight into the sustainability of OXLC NII and Core NII.
In my previous articles on CLO equity, I have argued that the precipitous drop in CLO equity prices was irrational and not based upon any realistic expectations of future leverage loan defaults. I predicted that even if the leveraged loans default rate would increase from the then current rate of less than 1% to a more normalized default rate of 2-3%, CLO Equity distributions would continue to be strong and expected returns (effective yield) would also stay in the range of 12-18%. According to S&P LCD, the leverage loan default rate is now just under 2% (still heavily concentrated in a few industries) and CLO performance still is quite strong.
The new issuance market is coming back to life (see chart 5). The rating agencies have downgraded very few CLO rated notes. Per Moody's, defaults held by CLOs as of April 30, 2016, were 59 bps. According to the WF CLO report:
- CLO junior OC test cushion is 401, and the interest diversion OC test cushion is 307 (more on the OC test below); and
- CLO equity distributions have averaged 4.81% of par on quarterly basis.
This is not to say that CLO equity investments are not still facing fairly strong headwinds; the leverage loan default rate will continue to increase and should start impacting other sectors than the Oil & Gas and Minerals & Mining sectors. According to Moody's CLO Global, Market Pulse reports all of the major CLO matrices are heading in the wrong direction. While all the matrices are within historical average and are fairly safe, the trend is quite troubling. In the remainder of my article, I will explain what these matrices mean and how they impact the CLO equity performance. In addition, I will compare OXLC's CLO equity portfolio matrices against the overall market and ECC.
OXLC provides investors with very valuable CLO level data, which provides investors with insight on how each CLO equity position will perform over time. As I will show in more detail below, the one beneficial thing about CLOs is that their underlying financial performance generally does not change in a short period of time, it generally takes time for a leveraged loan to get downgraded and then to default. CLO matrices are designed to give the investor insight into the performance of the CLO prior to defaults actually occurring as well as after (i.e., distressed priced loans and WARF and CCC buckets); all of these matrices combined provide the investor with a relatively accurate barometer for a CLO's future performance and ultimately of the sustainability of future cash distributions. Therefore, an investor in OXLC or ECC should have plenty of warning before a CLO equity stops distributing cash flows to their equity investors.
For confidentiality reasons, OXLC has decided to provide this information on a blind basis (without the names of the CLOs). See page 6 of its first-quarter 2016 investor presentation. This information plus the changes in each CLO's calculated effective yield provides the investor with truly valuable insight into the projected performance of each CLO investment.
CLO Structure Enhancements - CLOs have numerous structural safeguards that protect the interest of the senior rated notes, generally to the determent of the CLO equity (at least on the short term). The most powerful structural enhancement is the overcollateralization test (OC test), which is a calculation that takes the adjusted asset value of the collateral (not the fair value) over the par value of the notes. If the OC test (Sr. and Jr. OC Ratio Tests) is not satisfied, all cash flows that would normally go to the CLO equity are diverted and the funds are used to pay down the senior most rated notes.
The reinvestment OC test is similar to the OC test, but the test level is set a slightly tighter level than the Jr. OC test, 104% vs. 103% for example. Also, if the test is not satisfied, only 50% of the cash that would be distributable to equity is diverted and is used to purchase additional loans instead of paying down the senior notes. Absent the trapping of cash related to an OC test failure, CLO equity distributions should continue to be strong even if they incur defaults. It is my understanding that any CLO that is not failing its OC tests and is in its reinvestment period should be able to distribute at least 3.5-4.0% on a quarterly basis. Of OXLC's CLOs that are distributing less than 4% to equity, the vast majority of them are outside the reinvestment period (i.e. the deal is winding down) - (see page 9 of the first-quarter investor presentation).
An interesting note, even if a CLO is failing its Jr. OC test and the equity is not getting any distributions, OXLC will probably still accrue interest income and recognize taxable income on the investment. Just because a CLO fails an OC test for a period or two does not automatically mean that the CLO equity will have a negative return.
For all OC tests, loans are generally held at their par value unless they are deemed to be impaired in some manner. The two major adjustments to the par value of a loan are defaulted loans and excess CCC loans. There are other haircuts, but these are the two most important. Defaulted loans are held at the lower of fair value and assumed recovery rate (60%-17% for senior secured loans). CCC-rated assets in excess of the predetermined CCC limitation, usually 7.5%, are held at the fair value of the lowest-valued CCC assets. With the current low level of defaults, the excess CCC haircut is currently more impactful to the OC test than defaults.
Weighted Average Ratings Factor (WARF or WAR Factor) - Is a Moody's-based calculation that attempts to quantify the credit quality of a portfolio of leveraged loans. On average, broadly syndicated CLOs have an expected WARF of roughly 2,700-2,800. The higher a portfolio's WARF, the worse the credit quality of the portfolio. A high WARF score in and of itself does not tell the investor the entire story since some deals are simply structured with the expectation that the manager will invest in lower-rated loans. Middle market CLOs generally have an expected WARF in excess of 3,000. The CLO with the highest WAR factor in OXLC's portfolio is a middle market transaction.
As stated above, in the current relatively benign default environment, the CCC haircut is the most troublesome test and one that is most likely to cause an OC test or reinvestment OC test to fail. As loans get downgraded, more loans will be rated CCC and potentially cause the CCC test to fail causing the OC test to worsen.
I humbly ask the management of ECC and OXLC (assuming that they read my articles) to provide investors with the percentage of CCCs included in each CLO. This would provide investors with the best insight into which funds are struggling and who are potentially in danger of failing one of the OC tests.
OXLC's CLO Matrices
While OXLC does a pretty good job of providing its investors with detail portfolio information, it generally does it on a blind basis (without the names of the CLOs). ECC provides the information in a similar manner. In an effort to maximize the value of the CLO matrices, I matched each blind CLO with the actual CLO. This was an extremely laborious project, and I can't be certain my matching was 100% accurate. In an effort to respect OXLC's wish to keep this detailed information on a blind basis, I will not provide the reader with the information on a name basis, but I will break out the information between the top 10 investments and the remainder of the portfolio. In addition, I have included the WF CLO report averages and ECC's CLO equity portfolio information. Please see the chart below:
Comparison of OXLC's portfolio matrices
Please note that ECC does not provide its investors with an average CLO Equity distribution percentage, so I took its total CLO equity distributions for the first-quarter of $20.3 million (page 4 of its first-quarter investor presentation report) by the total par amount of its equity investment of $257.4 million (see page 2 of its semi-annual report dated March 31, 2016). The amount does appear to be high especially since it only had one investment make its first CLO equity distribution during the quarter. Also, please note that ECC's quarterly estimated taxable income in excess of its interest income is roughly $10 million, or $0.72 a share. When and if ECC ever decides to distribute some or all of the taxable income, shareholders should receive a very large distribution.
As the reader can see, the top 10 CLO equity investments are generally performing better than the remainder of the portfolio. Part of the reason for that is that Atrium XII and OZLM XIV, which were both recently issued, are included in the top 10. I do want to note that the one CLO that was failing its reinvestment OC test at March 31, 2016, is a significant investment but not in the top 10. Also, none of the investments that had large reductions in their effective yield (chart 4) were included in the top 10. However, one of my credit watch list name below (see chart 9 below) is in the top 10.
Overall, OXLC's portfolio is performing relatively well compared to Wells Fargo's (NYSE:WFC) and ECC's averages. All of the average WAR factors are right around 2,800. OXLC's increased by 49 points or just under 2% while ECC's increased by 77 points or just under 3% (this may be caused by new investments rather than credit deterioration). OXLC's OC test results are also mixed. Its average reinvestment OC cushion is above the Wells Fargo average; however, it showed significant deterioration over the quarter. The cushion fell by 128 bps or 27%. This is a very troubling trend and something investors should watch closely. In addition, OXLC's Jr. OC test cushion of 4.44% is above ECC's 4.27%; however, its cushions fell by 21 bps and 48 bps, respectively.
OXLC's average CLO equity distribution of 4.5% at March 31, 2016, is low and should increase to roughly 5.5% on an ongoing basis given the addition of Atrium XII CLO and OZLM XIV. The 5.5% compares favorably to the Wells Fargo average of 4.81%. If I calculated ECC's CLO equity distributions percentage correctly, its CLO equity investments are performing absolutely amazingly well. In fact, its performance is so strong I have a feeling that there is something wrong with my calculations. Hopefully, ECC's management will provide more insight into its CLO equity distributions. Either way, I will track it over time in an effort to determine if my calculation is correct.
As I discussed above, OXLC's WAEY fluctuates on a quarterly basis and provides the investor with an insight into management's most up-to-date thoughts on the long-term performance of a CLO equity investment. A significant drop in the fair value of a CLO equity investment, combined with a significant reduction in its effective yield, indicates that a CLO is encountering at least some stress. Based upon all the information provided by OXLC, on a blind and named basis, I have developed a watch list of CLO equity investments that may not be performing up to expectations. As the reader can see, the two Mountain Hawk deals appear to be performing the worst. The four deals on the watch list represent 10.0% and 4.5% of the portfolio on a cost and fair value basis. Also, not surprisingly, the one CLO that is failing its reinvestment OC test is on the list.
OXLC's Investment Watch List
Finally, one CLO matrix that is gaining prominence is the percentage of the underlying loans that are distressed and trading at a price under 80% of par. ECC provides the following chart, which is extremely helpful. Obviously, it would be more helpful if it would provide this information on a blind CLO basis. ECC's 5.8% of distressed assets is right in line with the 6.0% published in the WF CLO report.
It would be very helpful if OXLC's management would provide its investors with at least the same fair value information for its underlying loan exposure that ECC provides. Both OXLC and ECC provide their investors with the average fair value of the underlying loans in their CLOs, 90.70% and 91.97%, respectively. It would be my guess that both of their average loan price would increase by at least 200 bps, but that is only a guess. As stated above, it would even be more helpful if they could provide this information for each CLO on a blind basis.
As most readers know that my main goal in writing these articles is to provide my readers with information that will allow them to make the most informed decision possible. There is a huge gap between what information that is available to the professional CLO investor and an average retail investor. Intex is a computer service, much like Bloomberg, that has extremely detailed information on each CLO outstanding. The information includes each loan position, its rating and its fair value, as well as all the CLO matrices and tests. Unfortunately, I don't have access to Intex, but I have tried to use my detailed and professional knowledge of CLOs to help my readers bridge that gap. I hope I have succeeded in a least a little way.
Without taking too much credit, I have argued that CLOs are structurally strong vehicles and are built to withstand fairly high defaults and still perform. I believe the current state of the CLO market has proven that I was accurate in my prediction that the sell-off in CLOs was overblown and not tied to reality. I believe that CLO equity prices will continue to increase even as leverage loan defaults continue to rise. With the increase in default rates, there will be a divergence of performance among OXLC's portfolio. Some managers will perform exceedingly well while others will struggle. The one good thing about OXLC and ECC, they do have access to the best CLO managers and their portfolios are quite diversified by vintage and by managers.
As most readers know, I don't like to give buy/sell ratings on stocks, for the one reason, I am not the world's best stock picker. I do like to provide the reader with the information to make the most informed decision. However, I do have a couple of thoughts on OXLC.
First, readers must know that OXLC's management has decided to include a significant portion of return of capital (ROC) in its dividend. So when an investor goes to Yahoo or Google Finance and sees that OXLC is yielding roughly 25%, they must know that almost half of that is ROC. Its dividend yield on a NII basis is roughly 15% while ECC's is roughly 14%. If an investor is holding their OXLC's shares in a tax-advantaged account, there is nothing inherently wrong with returning ROC to its shareholders. However, it does cause the company to need to issue share from time to time.
Second, OXLC's portfolio is generally in good shape and should continue to pump out cash flows for the next few quarters (at least). The company's problems with its asset coverage test should be behind it as CLO equity valuations improve and/or it repurchases additional preferred shares. I believe its dividend should be safe for at least the remainder of the year.
Finally, an investor can't forget that OXLC just recently sold $55 million of CLO investments at or very near to the bottom of the market and incurred $34 million or $1.13 a share of realized losses. No matter how you spin it, that's not good.
Overall, I guess my take on OXLC is similar to my old and wise colleague at S&P, Steve Bavaria, who has stated, ECC is a core part of portfolio, like ARCC, that I bought and don't worry too much about. I consider OXLC a good trading stock, I own some that I bought around $8 and I will keep for a while. I like the cash flows and at least for I don't see any major troubles brewing. Good luck!!!
Disclosure: I am/we are long ECC, OXLC.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.