What are collateralized loan obligations (CLOs)?
CLOs are essentially a pool of bank loans that were securitized. In other words, a structuring agent (let's say a structured finance desk at a large bank) would collect a large portfolio of senior secured bank loans, put it in a special purpose vehicle (SPV) and have it issue various securities backed by the underlying collateral pool. Those issued securities have specific assigned order of cash flow distribution and loss absorption related to the underlying pool of loans.
Effectively, a structuring agent takes a pool of typically non-investment grade bank loans (let's say with BB equivalent rating) and then through assigning different order of cash flow and risk distribution develops a wide spectrum of securities: from AAA all the way to equity (aka residual tranche). AAA tranche is the first one to get cash flows generated by an underlying pool of loans, while equity is the last one (after all debt tranches are fully repaid). However, not all of the loans will be paid back. Such loan losses are absorbed starting from lowest tranche, i.e. equity tranche. In other words, equity tranche of CLO is the last in line to receive cash and first to absorb losses.
Typical CLO Structure (Source: Priority Income Fund offering document)
Naturally, you would think that regulators want sponsors of securitization vehicles to have some skin in the game (especially, given the experience of 2008). Indeed, Section 941 of the Dodd-Frank Wall Street Reform (and some other regulations) requires the seller, sponsor or securitizer of a securitization vehicle to retain no less than five percent of the credit risk in assets it sells into a securitization and prohibiting such securitizer from directly or indirectly hedging or otherwise transferring the retained credit risk. However, when it comes to CLOs, it is not as straightforward due to some rulings and uncertainty of how it is to be implemented in practice. Furthermore, 5% stake might not be sufficient enough to deter poor behavior given how lucrative could be putting together such structures.
What is OXLC?
Oxford Lane Capital Corp. (OXLC) is a closed-end fund that invests predominantly in equity tranche of CLOs. In the July 2018 article, "Trading Softly" (Seeking Alpha contributor) made a case to buying OXLC based on the following key points: 1) an attractive dividend yield of 15-17%, and 2) balancing impact of prime rate increases (on one hand heavily indebted companies might struggle with increasing rates; on another hand, CLO would experience increasing interest income given the floating nature of underlying loans). In December 2017 article, "Stanford Chemist" (Seeking Alpha contributor) provided more reasons that make OXLC (and ECC) an attractive proposition: historically superior risk-return profile compared to debt instruments, low interest rate sensitivity, and stronger credit quality (bank loans are senior secured).
How does OXLC achieve such high dividend yield? Well, it does so by investing in the riskiest portion of CLO; so-called equity or residual tranche. OXLC had 98% allocation to equity tranches and 2% in debt tranches. As of Q3'18, equity tranche had a cash distribution yield of 20.7% annualized.
How come equity tranches earn such high yield? Well, that is due to high leverage ratio (~10x) of the CLO structure and credit risk. To give you some additional color on this, let's refer to another Seeking Alpha contributor, Steven Bavaria, who provides an interesting observation stating that OXLC offers an equity-like return (yields in the low to mid-teens), but have largely traded the equity risk for credit risk. In another article, Steven makes another observation that "CLO is like a virtual bank, and investors in the equity of CLOs are like shareholders of the virtual bank". So Steven is implying that OXLC is effectively a virtual bank. Do you know what happens to banks during a recession?
That said, investing in shares of OXLC would not be exactly comparable to owning a bank stock. On one hand, similar to bank equity exposure, you get leveraged exposure to credit risk by buying OXLC. On another hand, OXLC has a 2/20 fee structure that is more of a hedge fund fee/incentive structure than the one you expected from your community bank. Per the semi-annual report of OXLC, charges 2% base management fee and 20% incentive fee with 7% hurdle rate (with catch up provision).
A quick note on the impact of the interest rate increase on OXLC
There are two components of CLO that would be relevant when we talk about interest rates: 1) underlying loan portfolio, and 2) capital structure of CLO itself.
Underlying portfolio (item #1) will benefit from an increase in interest rates thanks to the floating nature of bank loans. That said, as rates continue to increase, heavily indebted borrowers might start experiencing distress that can increase credit losses. In other words, in the early parts of increasing rate environment, the portfolio would benefit from rate increase.
An increase in LIBOR would increase the CLO vehicles' financing costs. However, most of the collateral positions within the CLO investments are yielding rates below LIBOR (i.e., have LIBOR floors below LIBOR currently) and will start yielding more as LIBOR increases.
What makes OXLC a "short" candidate?
Hm, let's put fee/incentive structure aside and focus on the risks. Here's the list of some of the risks and how it was captured in OXLC's semi-annual report:
- Capital market disruption happens (last big one was in 2008), and during such periods credit, spreads tend to increase, which hurts any credit product, e.g. corporate bonds, high-yield bonds, bank loans, etc. Leveraged instruments like equity tranche of CLO tend to perform especially poorly during such period.
The U.S. capital markets have experienced periods of volatility and disruption. Disruptions in the capital markets tend to increase the spread between the yields realized on risk-free and higher risk securities, resulting in illiquidity in parts of the capital markets.
2. CLOs have additional structural requirements that make equity tranche especially prone to negative impacts of market disruptions: "waterfall" ensures that equity is the first one to absorb losses, over-collateralization and interest coverage requirements might further divert cash flow away from equity tranche.
The failure by a CLO vehicle in which we invest to satisfy certain financial covenants, including with respect to adequate collateralization and/or interest coverage tests, could lead to a reduction in its payments to us.
3. An equity tranche of CLOs is illiquid security (mostly privately placed):
The interests the Fund has acquired in CLO vehicles are generally thinly traded or have only a limited trading market. CLO vehicles are typically privately offered and sold, even in the secondary market. As a result, investments in CLO vehicles may be characterized as illiquid securities.
4. High embedded costs (remember 2/20 that we've discussed earlier):
Let's put items #2-#4 aside, and focus on item #1, i.e. impact of market disruptions on CLOs.
What could be the impact of the loan losses on equity tranche of CLO?
Here is the rough net return math for equity tranche of CLO (this is very simplified view only for demonstration purposes):
Net return = ("excess spread" - "credit losses") x "leverage"
Here, the excess spread is the difference between the average interest coupon earned on the underlying pool of loans (asset return) and cost of liability in CLO structure. In the case of OXLC, this math would indicate that excess spread, net of losses, on equity tranche of CLOs is at ~2% (i.e., 20% cash distribution rate divided by 10x leverage).
If the underlying portfolio coupon stays stable (note that OXLC has some portfolio churn) and experiences 1% increase from current levels (i.e. if the current rate is 0.5%, then this scenario would assume 1.5% loss rate), net return would be:
Net return = (2% - 1%) x 10 = 10%
In other words, the increase in loan loss by 1% results in 10% deterioration of net return due to leverage. So, now the question is whether 1% is too small or too big of a number. We will address it by considering the experience of 2008.
What would be the impact of the 2008-like event on OXLC?
Wouldn't it be nice to have OXLC around during the last recession and see how it performed? Unfortunately, neither OXLC nor ECC was around. Given this, let's use math offered by Steven, with some adjustments that we made, to estimate what would be the impact of the 2008-like event on OXLC. Quick reminder: all we need is to figure out what was the peak loan loss in 2008 multiply it by 10, which is the average leverage of the portfolio as noted by OXLC quarterly publication (actual leverage is 10.5x, but for the sake of round numbers, I'm using 10x).
So, let's take a look at the graph below from St. Louis Fed's website:
That peak in Q1'2010 was 3.12% of net loan losses. Note that four years prior to peak, i.e., in Q1'2006, the same stood at 0.35%. Another interesting observation: that loan losses tend to start to increase a few years prior to recessions.
So, let's run the numbers: 3.12% x 10 (leverage) = 31.2% or let's just say ~30%.
But, wait, that's not all. Remember, OXLC is CEF (i.e., not ETF). What does that mean? It means that OXLC might as well be trading at discount to its NAV. We have estimated that impact on underlying could easily be 30% (which might end up being far too low of an estimate vs. might actually happen during next credit crunch), but CEFs, unlike ETFs, can trade at premium or discount (see my previous article if you would like to learn more about this aspect).
How much could be discount?
In the case of OXLC at times discount was at around 20% as you can see in the graph below, but it can also trade at ~50% premium as well. For purposes of our calculation, however, the question should actually be how much discount can change in a matter of short period of time. If you bought OXLC at 20% discount to NAV and sold at the same 20% discount a few months later, your investment performance would mostly reflect the change in NAV and dividend income during the period.
For our calculation, let's assume one year time. One can argue that as we approach economic slowdown, OXLC would likely already trade at discount in anticipation of a recession. So, one might not observe ~70 point change in pricing: from ~60% premium to ~10% discount as it happened between Q3'16 and Q2'17. That said, in the last one quarter alone, we've observed going from a 10% premium to 10% discount, which is a 20% move.
So, let's put this together. One can argue that in the 2008-like event, OXLC is likely to experience a 30% decrease in NAV and 20% change in discount. For the ease of demonstration, let's assume that just prior to the event, OXLC was trading at $10 NAV and $10 share price (i.e., zero discount). At the time of the 2008-like event, $10 NAV would lose 30% and become $7. Now let's apply a 20% discount on this NAV, and we get a price of $5.6 per share. That is 44% price deterioration (from $10 pre-event to $5.6 post)!
So, OXLC might as well exhibit SPY-like downside during the next downturn. Or, perhaps, it might experience even higher drawdowns. At the end of the day, didn't we say that buying OXLC is comparable to buying bank shares? And what happens during the recession to bank shares? Perhaps, from this perspective, 44% might end up being a very low estimate of potential downside.
As noted in my previous article, ultimately, you would need to decide yourself when and what to short. It might be very early to short OXLC or any other credit product at this stage. But as we get closer and closer to peak, the proposition to short credit products would become more and more attractive. Personally, I will put a relatively small short position (that would likely represent less than 0.5% of my investable assets) and keep an eye on how things develop. If the market starts moving in my favor, I am going to increase my position. That said, this is not investment advice! Please consult your judgment and/or investment advisor (whichever you prefer).
I am not claiming that I have a crystal ball. Instead, I'm offering my views on how OXLC is likely to perform during the next downturn. Shorting OXLC today comes with 1) high costs for carrying this short position (OXLC pays high dividend yield, and fees to short can be significant even for a relatively small short position), and 2) risk that market downside might not happen anytime soon. Perhaps, one should wait until they are sure that credit window is closed shut. Or, perhaps, it would be too late to short "after the fact".
Disclosure: I/we have no positions in any stocks mentioned, but may initiate a short position in OXLC over the next 72 hours. 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.