Credit Risk, Equity Returns: Interpreting Oxford Lane Capital's Financial Results

| About: Oxford Lane (OXLC)

Summary

Closed end funds like OXLC and ECC that hold collateralized loan obligations can be a challenge to analyze, but their (hopefully) steady yields in the teens make it worth it.

Taxable (cash) earnings can differ substantially from GAAP earnings, so that a fund may appear to have plenty of cash flow to pay distributions while its GAAP-derived NAV is dropping.

Are these accounting idiosyncrasies and one-time events, or trends actually worth worrying about? Or a bit of both?.

Beats the heck out of me! (Just kidding...sort of).

For the intrepid band of Seeking Alpha contributors and readers who own closed end fund Oxford Lane Capital Corp. (NASDAQ:OXLC), the periodic task of interpreting the fund's quarterly earnings announcement and investor presentation to ascertain whether OXLC's generous 14% dividend is sustainable is critical to being able to sleep well for the next three months. Part of the challenge is understanding the implications of the GAAP earnings being a modest 7 cents a share for the quarter (driven by mark-to-market paper losses) while the taxable earnings (on which distributions are based) was a more robust 53 cents per share.

So here we go.

First some background: OXLC was the first closed end fund to specialize in collateralized loan obligations ("CLOs"), an asset class previously restricted to pension funds, endowments, hedge funds, and other institutional investors. OXLC opened its doors in 2011 and first came to the attention of the author early in 2013. I have written several articles about it (like this one) and mention it in most of my "savvy senior" portfolio review articles (most recent). Since buying OXLC in April 2013, I have enjoyed the 14 to 15% annual dividend, although I have given back a bit over 2% of that, annualized, in price erosion, for a net annual return of about 12%. To me, that is an "equity" return, and I am happy to get it. Getting it in cash dividends is particularly attractive, since my actively managed portfolio is exclusively in IRAs, where I focus on compounding dividends tax-deferred while trying not to concern myself much with how the market values my portfolio from week to week or month to month.

As mentioned in a recent article, the advent of Eagle Point Credit (NYSE:ECC) means there are now two closed end funds that allow retail investors access to the CLO market. I recently bought some ECC, which has already declared its first quarterly dividend to be paid in January. The yield is about 10% at ECC's current price, which has risen since the dividend was first announced. At that time it was closer to 12%, the target yield the fund's management has estimated they expect the fund to meet going forward. I am happy to see ECC come on the scene for several reasons: (1) It is nice to have another CLO-focused closed end fund, rather than have all our CLO eggs in one basket; (2) It will provide an additional source of transparency about the asset class, helping us to understand both funds better; and (3) it is further evidence that the CLO market, in its matured, post credit-crisis state, may continue to widen its investor reach to the retail market.

OXLC and ECC buy equity (some debt too, but mostly equity) in CLOs. Think of a CLO as a virtual bank. They leverage their equity at about 10 to 1, typically borrowing at about 1.5%, and use the proceeds to buy senior, secured corporate loans syndicated by major banks (JPMorgan, Citibank, BofA, etc.) with floating-rate coupons based on a spread - typically 300-600 basis points - over 3-month LIBOR, depending on credit and structure, resulting in a coupon of about 4 to 7% in today's market. The CLO's average spread between its assets and liabilities can therefore be somewhere from 3% to 5% or more, which leveraged 10 times, looks like a great return for the equity: 30-40% and more. But before that gross spread becomes net, and payable to equity holders, you have to deduct a whole range of expenses, like credit losses, adverse changes in interest rates, costs of loan prepayments, or other items that can reduce the profit margin on the portfolio. And at 10 times leverage, bad news, so to speak, travels fast through your portfolio.

Despite the risks, CLOs have been a pretty stable buy-and-hold asset for long-term income investors willing to ignore market volatility. The CLOs created in the 2000-2007 era, despite horrific declines in the price of both the CLO equity itself and the underlying loans held by the CLOs during the 2008 meltdown, virtually all came through with positive returns for steadfast investors who hung on tight and did not sell out. In that, their experience mirrored that of investors in the equity and high yield bond markets who rode out the storm. (Other asset classes, especially securitized vehicles that bought poorly underwritten home mortgages and home equity loans, got creamed and lost their investors billions. It did not help the reputation of the CLO market that the lousy real estate deals were packaged into similarly sounding vehicles called "collateralized DEBT obligations" or "CDOs." What a difference a single letter can make!)

It's the distribution, stupid!

The key to being a happy CLO or CLO fund investor is gaining assurance the CLOs will continue to pump out the cash required to pay distributions. This is a tricky exercise because the definition of earnings for a CLO (or a fund that owns CLOs) is a bit complicated. [Caveat: The author spent many years working in the credit markets and has a good grasp of corporate loans - i.e. the loans CLOs hold. But the market for securitizing those loans into special purpose vehicles (e.g. CLOs) is a somewhat arcane one. So I am only a half step or so ahead of many readers in figuring this asset class out and trying to explain it. But I think it's worth the effort because CEFs like OXLC and ECC offer retail investors an opportunity to get into a high-yielding asset class previously restricted to institutional investors, if we can figure it out and become comfortable with it.]

Having said that, here's how I think it works.

CLO investors are buying pools of loans that, in theory, will provide them with cash flows over a 12-year period (the typical life of a CLO) far in excess of the average cost of the liabilities funding the CLO. But there are lots of factors that can arise that reduce the actual funds flow from what "in theory" they should be. CLO investors, like our two funds OXLC and ECC, create sophisticated models where they project what they believe the cash flows will really be, after providing for various contingencies. Those modeled cash flows are used to project and accrue GAAP earnings from the CLOs in the funds' portfolios. However, as I understand it, investment companies can only charge off certain expenses for tax purposes if and when they are actually incurred. Distributions to shareholders are based on taxable income, not GAAP income. But GAAP income is used to determine changes in NAV. This is why we have seen, in OXLC, NAV drop over the last three quarters because cash/taxable income has exceeded GAAP income. In other words, from a GAAP standpoint - accruing for expenses that were not incurred on a cash basis - the distribution does not appear to have been covered. But from a taxable income viewpoint, there was sufficient cash flow/taxable income to cover it.

The difference between the two - taxable income and GAAP - can be striking. For example, in its latest quarterly report, OXLC said the "effective yield" of its total investment portfolio, on a GAAP basis, was 11.3%. Meanwhile, the weighted average taxable yield of "income producing" investments was reported to be 24.3%. In other words, 24.3% is the actual cash yield that OXLC is receiving, on those assets that are actually earning income. Unraveling the difference between those two numbers is important to understanding how CLOs work. The 11.3% yield number includes all sorts of assumptions and provisions for various costs and is intended (I think) to represent the yield OXLC would earn on its portfolio over the course of its life if all the assumptions that are baked into its model turn out to be accurate. To the extent that the model assumptions turn out to be overly conservative, then the return is likely to be closer to the cash/taxable yield, although it may not be as high as 24%.

What are some of the factors that are baked into the GAAP model that can ultimately reduce the CLO investor's cash flow?

· An obvious one is credit expense. An investor may assume that 3% of its loans will default annually (these are non-investment grade loans so some defaults are to be expected), but since they are all senor and secured, the recoveries tend to be high (about 75%), so the loss rate would be about 25% of those that default, giving a credit loss expense of 3% times 25%, or 75 basis points (0.75%) per year. Leveraged 10 to 1 that would be a 7.5% hit to the CLO investor's earnings, and would be built into the model that determines GAAP income. In recent years corporate defaults have been unusually light, even negligible for many CLOs, so this factor alone could account for a 7% or so difference between taxable and GAAP yields.

· The lag before new CLO equity begins to make payments. There is often a one or two-quarter lag before equity on newly issued CLOs starts to receive cash payments. This can make a big difference. In OXLC's 11/12/2014 presentation, it reports that of its approximately $300 million of CLO equity investments, about 47%, or $144 million, was newly acquired CLO equity that would not start generating income until the fourth quarter of this year (most of it) or in some cases the first quarter of 2015. But the weighted average taxable yield of the 53% of CLO equity investments that was generating cash income in the 9/30/2014 quarter was 26%. That's a lot of cash generation waiting in the wings, that will presumably be available to support the distribution next quarter and beyond. I presume that this periodic delay in earning cash income on newly acquired assets has been factored into the model, and is another contributor to the difference between reported GAAP earnings and actual cash flows.

· The eventual rise in interest rates. At some point interest rates will rise, bringing LIBOR up with it and that will immediately raise the cost of all CLO liabilities, which are based on LIBOR. "Not so fast," you say, "aren't the CLO's loans also based on LIBOR, and won't their coupons rise too?" The answer is, "Not quite." Ever since short-term interest rates plummeted to near zero, corporate loans have been written with LIBOR "floors," of 1% or 1.5%. This means even if LIBOR is 0.25%, the loan's base rate (to which its credit spread is added) is 1% or whatever floor is specified on that loan contract. That's been great for lenders in recent years when rates have been below the floor, but it also means when rates start to move up, the rates they charge their borrowers won't move up for the first 75 or 100 basis points or so, until the actual 3-month LIBOR rate reaches the floor that the loan has been using. Meanwhile the CLO's own liabilities are based on actual, "unfloored" LIBOR, so they will see the cost of their liabilities move up immediately. Leveraged at 10 to 1, that means a potential 7% or more hit to their yields. I am told that a provision for this eventual cost is also factored into CLO investors' earnings models.

· Finally, there is the "run off" factor. CLO equity benefits from leverage that is much cheaper than what the CLO earns on its assets. The reason their liabilities are so cheap is because CLO senior creditors get paid off first when the loans are repaid, and the CLO junior debt and equity holders get paid off last. If a CLO winds down gradually, there is a period when a lot of the cheap, triple-A rated liabilities have been paid off but the more expensive junior liabilities are still outstanding. Since those junior liabilities may carry coupons as high or higher than the loans the CLO owns, the cash flow to equity could flatten out or even turn negative as the portfolio winds down. Therefore CLO equity holders have an incentive to accelerate the wind-down of their vehicles once they reach the end of the reinvestment period (when repaid loan principal can be reinvested), rather than have a long gradual wind-down. Here again the fund presumably has recognized this issue and incorporated various assumptions about it in its model (as well as various strategies for dealing with it in its planning.) One obvious assumption is that the equity owners of a CLO will get together and move to accelerate the wind-down of the CLO. This is easier if the fund holds majority interests in the CLOs it owns (ECC says it does. I don't know if OXLC does.) Even if nobody owns a majority of a particular CLO's equity, it probably still makes sense for the equity owners to band together to agree to accelerate their CLO.

Conclusion: Anyone whose eyes haven't glazed over and is still reading this article will notice by now that CLOs have a lot of moving parts. In one sense, it is simple. If all the loans pay off as scheduled (i.e. no credit losses), and interest rates don't change, and the CLO manages to buy seasoned CLO equity that is already making equity payments, or buys new equity at a discount from par that offsets the timing delay, and manages to accelerate the winding down of its investments so that its spreads remain largely intact, and everything else goes just perfectly, then the return could be up around 30% or so. However, in the real world some borrowers do default, interest rates move against you and other bad things happen. We hope and expect that the fund managers have thought of all those things (and more!) and incorporated them appropriately in their models, and that the distribution policies they have embarked on are consistent with and supported by their models.

Personally, I believe that is management's intention, to incorporate if not a worst case, than at least a serious real-world case in the models, and plan distributions based on those. If they have done that, then any surprises are more likely to be positive ones - special one-time payouts (like we got earlier this year) because taxable earnings turn out to be better than their model assumptions. As substantial investors in their own funds, that would be a rational course, rather than to have distributions fluctuate, which would scare off investors seeking consistent income.

Time will tell. In the meantime I will continue to be an owner of both funds. Let's also hope that fund management will see that it is in their own interest to be as transparent about all of this as they can. I think it is in both OXLC's and ECC's interest to help educate investors about their operating models, so we can all become "high conviction owners," and not just "low conviction yield chasers" who are more likely to sell and run at the first sign of trouble.

Disclosure: The author is long ECC, OXLC.

The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it (other than from Seeking Alpha). The author has no business relationship with any company whose stock is mentioned in this article.

About this article:

Expand
Author payment: $35 + $0.01/page view. Authors of PRO articles receive a minimum guaranteed payment of $150-500.
Tagged: , , , Closed-End Fund - Debt
Want to share your opinion on this article? Add a comment.
Disagree with this article? .
To report a factual error in this article, click here