PSEC trades at a premium to competitors on both price-to-earnings and price-to-book measures. When I first submitted this idea for VIC membership, the stock traded at a premium to book value. Today, the price is at a slight 16% discount to book value. Nonetheless, I believe the stock is still overvalued. Management has a history of significant value distruction, and the following factors suggest that fair value is still well below the market price of the stock:
1. Management incentives are to deploy capital at all costs and inflate asset values
2. Portfolio-level return data suggest that fair values are overstated
3. Recent dividends in excess of net investment income are unsustainable and will be cut
4. Other signs of accounting manipulation - internally inconsistent marks and general obfuscation
Management incentives are misaligned
PSEC management has misaligned incentives that cause aggressive lending and overstatement of assets. Management earns 2% on all assets under management, 20% of net investment income, and 20% of capital gains (subject to 7% annualized hurdle). Interest income includes significant portions of non-cash income, including Pay In Kind (PIK) interest and accretion of loan discount. By separating capital gains from investment income, PSEC ensures high fees as long as it makes loans with high interest payments, regardless of whether those loans perform in the medium term. Thus, incentive fees are protected even during periods when investors lose capital. For example, during 2008 and 2009, PSEC managers took home $11.3M and $14.8M in incentive-based fees, even as the company's stock price cratered.
Due to this compensation structure, management's primary incentive is to deploy as much capital as possible at high interest rates, regardless of investment risks. This incentive likely accounts for the fact that total financial assets have swelled from $94M in 2005 to $1,308M today. Consequently, old, bad investments are easily dwarfed by the much larger base of fresh loans, all generating high (and temporary) investment income. PSEC issues hundreds of millions of dollars yearly in dilutive common stock to fund this growth (and to cover persistent shortfalls in cash from operations). Moreover, 2% and 20% are not inherently wrong incentives, but the combination of an unsophisticated investor base and lack of actual performance metrics given the constant inflow of capital suggests that rewarding managers exclusively for expanding the capital base is inappropriate at best and reckless at worst.
Furthermore, these incentives also create pressure to inflate asset values. In FY 2010, PSEC acquired Patriot Capital Funding, and recognized an immediate gain on the assets of $7.7M. Notably, PSEC recognized the gain as "other income" within "investment income," rather than as an unrealized capital gain. In my view, it would have been more appropriate to classify the transaction as an unrealized capital gain, but this would have delayed (and jeopardized) any incentive fee. Through this one transaction and immediate write-up, PSEC managers increased their 2010 incentive fee by ~$1.5M. It is fishy in the first place that PSEC should purchase Patriot's assets for one price and then recognize an immediate accounting gain. However the circumstances under which PSEC not only recognized the gain but also classified it in a way that increased the bonus pot by $1.5M are especially suspect.
Portfolio-level data suggests that fair values are overstated
Analysis of portfolio level data also suggest that fair values are overstated. Since 2007, average yield on fixed-income assets has been between 14.3% and 16.8%, very high yields. For high-yielding debt, a significant portion of the yield is attributable to expected default losses. Furthermore, it is well-documented that default probabilities for high-yield debt accelerate in years 3 and 4, with default in years 1 and 2 comparatively unlikely. This is especially true for PIK loans or loans with high original bond discounts, as much of the yield is non-cash (PIK and amortization of discount). Given these circumstances, PSECshould account for some write-downs of all high-yielding loans within the first 1-2 years, even before the loans default. This is standard practice for retail banks, and should especially be employed by PSECgiven PSEC's high-yielding portfolio.
However, PSEC has a tendency to delay write-downs until absolutely necessary. For example, during FY 2010, PSEC took $36M of write-downs, of which $22M were accounted for loans in non-accrual status. Of the $36M, all loans had been originated in 2006, 2007, or 2008, with no write-downs on any loans in 2009 and 2010. Meanwhile, PSEC was recognizing interest income and writing up new loans through PIK interest and amortized loan discount. No loans originated in 2009 or 2010 had any write-downs associated, despite the fact that those loans had very high yields in 2009 and 2010, and that these yields were associated with significant default probabilities expected in future periods.
The pattern of only recognizing default once loan reach non-accrual status is repeated in other periods. Moreover, it creates a situation where asset values are perpetually overstated, especially during periods of asset growth, as the first 12 - 24 months of a loan is typically too early for credit events but early-enough to recognize the extremely high yields in investment income. With fair value investments having grown at 68% over the past 18 months (through December 31st 2010), fair values are significantly overstated, and a correction will come as the loans mature and some experience default.
Recent dividends in excess of net investment income are unsustainable and will be cut
PSEC's dividend is currently unsustainably high. By regulation, PSEC must pay out at least 90% of net investment income plus realized gains as dividends. However PSEC has gone beyond this metric recently, paying out on average 25% more dividends than net investment income plus realized capital gains. In the most recent quarter, PSEC distributed 34% more dividends than net investment income plus realized gains.
Notably, these dividends do not make economic sense, because PSEC typically raises more equity in a period than they pay-out. Consequently, the dividends result in a taxable event for shareholders even as PSEC raises fresh capital and dilutes the same shareholders. This policy will destroy value over time (although it sure is friendly for Uncle Sam). Consequently, it is reasonable to expect PSEC's dividend payout ratio to stabilize at or below 1x net investment income over the long term. Given that the current ratio is significantly higher than 1x, the dividend will likely fall.
Other signs of accounting manipulation - internally inconsistent markets and general obfuscation
PSEC's internally inconsistent marks also suggest that write-downs are only taken in extreme circumstances. In many cases, PSEC holds investments at different levels in the capital structure. It stands to reason that if one tranche - for example the subordinated secured debt - takes a write-down, other tranches would also likely have impaired fair value. However in many cases PSEC takes write-downs on one tranche without taking write-downs on other tranches. This policy may be consistent with PSEC's commitment to only taking write-downs once a credit event is highly likely, but it is not consistent with the concept of fair value, which should reflect both likely events and less likely events.
General obfuscation is a further worrying sign that PSEC's accounting is not forthcoming. For example, PSEC changed the names and corporate entities of Appalachian Energy Holdings ("AEH") and Conquest Cherokee, Inc, making it more difficult for analysts to track these investments.
PSEC's investments in AEH and Cherokee began to sour during FY 2009, when both companies stopped making interest payments. By October 2009, PSEC reorganized the debts of Cherokee, Inc into a new firm called Coalbed, Inc. Coalbed subsequently failed to make its first interest payment. In January 2010, the assets of Coalbed and AEH were combined into a new company, Manx Energy, which was responsible for both firms' debts, as well as a new loan of $2.8 million. In its new form, Manx Energy also failed to meets its initial debt service.
In the end, this has the appearance of a shell game - moving bad assets from one corporation to another and recapitalizing. The side benefit of reorganizing the debts of a company under a new firm is that within the 10-K it is not required to mention the total length or amount of the write down on the original loan, only the length of non-accrual and the amount since the reorganization. These corporate structures and disclosures make it more difficult for analysts to track PSEC's investment portfolio, and provide circumstantial evidence that PSEC may be guilty of accounting manipulation.
Summary and valuation
In summary, all indications suggest that a significant discount to book value should be taken to reflect PSEC's aggressive accounting as well as management's history of destroying value. It is hard to know exactly how much PSEC's underlying assets are actually worth, but the current price represents an attractive risk-reward to short the stock. At most, PSEC is worth 1x book value, just 16% above current prices. The downside of the stock, however, is significantly greater -- the vast expansion of capital may be hiding deep problems.
Investment risks and mitigants
PSEC's portfolio is significantly exposed to economic cycles, and sustained economic recovery would probably increase fair values.
PSEC has exposure to equity assets as well as debt assets, and appreciation of PSEC's equity portfolio is also a risk. However this risk is limited due to the relative size of PSEC's equity portfolio. Preferred shares, common equity, and other equity-like instruments account for just 11.1% of PSEC's investment portfolio.
If PSEC continues to expand its capital base, it may be able to hide problems of underperforming loans. Growth from ~$100M to over ~$1B in financial assets since 2005 has largely achieved this task, as the high investment income from new, high volume loans has shielded focus from the capital losses of older loans. However it is hard to imagine that PSEC can continue this growth trajectory (CAGR 55%) on top of $1B of capital. Instead, it is more likely that PSEC's portfolio will mature, with more seasoned loans beginning to default and exposing problems with PSEC's asset values.
The main catalyst is the maturation of PSEC's portfolio, which will force PSEC to recognize more and more capital losses over time.
Another possible catalyst would be a dividend cut. Current dividends are 100% funded by new share issues, and this constant inflow of cash from new investors is unsustainable. Furthermore, dividends significantly exceed net income. The fact that PSEC pays out dividends in excess of net income despite the fact that PSEC also raises cash every year is bizarre, as well as tax inefficient. Dividend yields may be reduced in the future.