Description
Allied Capital (ALD) has been written up twice before on VIC as a long. This is a long idea too go long the puts. This trade is very timely and has excellent risk reward. I will discuss the specific catalysts below. ALD is certainly not an unknown short idea but I believe many shorts have grown frustrated with the position and covered in disgust, never to revisit again. The short thesis was laid out by David Einhorn at a charity dinner a year and a half ago. You're probably asking yourself, 'What in the world can this guy pgu103 add?' Well, please keep reading.
Overview
I would suggest reading the previous two VIC write-ups. One of them actually addresses some of the short arguments raised by Einhorn point by point. In general, ALD is a business development company that provides mezzanine and equity financing to private companies. In addition, ALD manages a large portfolio of CMBS.
The biggest issue for the shorts is that ALD has discretion on how it values its various private debt and equity investments. In my experience, when management has discretion to make assumptions in determining its P&L, most of the time, the assumptions prove aggressive and investors later learn that the company wasn¡¦t making the kind of money that they thought it was. See the subprime lenders and their gain on sale assumptions or Enron and Dynegy's assumptions about the mark-to-market value of their derivative books for some recent examples. Wait a minute, just because the other guy's were aggressive with their accounting doesn't mean that ALD's investment valuations are aggressive. That is true. However, I am merely pointing out that when the fox is left to guard the hen house, trouble is more apt to happen. Human nature, I guess.
Another big picture issue with ALD is that it is impossible to tell how much of the income they are reporting is simply interest income that is being paid by the borrower with funds just lent by ALD. Certainly a large part of ALD's fee income is paid directly out of monies just lent to the borrower. Thus it is impossible to gauge the quality of ALD's earnings. Further, PIK income is almost 30% of ALD's operating income.
Key Questions
So here are the key questions to ask and answer if you have a bearish view on ALD:
- How can you be certain that ALD has inflated the valuations of its investments in mezzanine debt and equity?
- Why short ALD now?
- Isn't ALD expensive to short given that you have to pay the dividend and also have to pay to borrow the stock?
ALD Portfolio Valuations
First, while you can never be certain of a valuation until an actual exchange between willing third-parties, there is clear and convincing evidence that ALD will need to write down its two biggest private finance investments in the very near future. I will focus on these two investments simply because they are the highest profile investments in ALD's private finance portfolio. Also, you could make a career out of going through every single investment and gathering data on it. I have analyzed more than these two investments in detail but won¡¦t discuss them in this write-up for the sake of time and space. In addition, I am not even going to comment on the CMBS portfolio. Part of my thesis is that if you can prove the two biggest investments are mis-valued, the odds are high that a number of the smaller ones are too. Call it culture issues or bad modeling or whatever you want.
Business Loan Express Valuation
The first private finance investment I will discuss is Business Loan Express or BLX. BLX is an SBA lender. ALD's cost basis in its BLX debt and equity investments is $258 million and the current reported FMV is $345 million. The BLX investment (at FMV) represents 12% of ALD¡¦s assets, 20% of the private finance portfolio and 20% of NAV. It is probably helpful to note the following changes in cost, FMV and unrealized gain (in $millions) for the BLX investment.
Date Cost FMV Unrealized gain
12/31/01 212 228 16
3/31/02 214 230 16
6/30/02 217 252 35
9/30/02 219 254 35
12/31/02 221 257 36
3/31/03 226 308 82
6/30/03 265 357 92
9/30/03 258 345 87
The change in unrealized gain YTD (and YoY) is $51 million or a 20% increase. This increase in unrealized gain has accounted for 45% of the increase in ALD's NAV for the current year. (As an aside, notice the jump in valuation in the 3/03 quarter, shortly after the audit was complete and the auditors not being due back for another year.) What is driving this increase in unrealized gain? Great question.
As you can tell from the table above, ALD has invested an additional $39 million in BLX over the 12 months ended 9/30/03. During the same time period, BLX's EBITM (earnings before interest, taxes and management fees) DECLINED from $47 million to $42 million (as adjusted for one-time items), or an 11% decrease. Accoringly, ALD reported a 20% increase in its BLX unrealized gain for a period of time when BLX's earnings were decreasing. Of note, originations declined sequentially in the most recent quarter as BLX seeks to diversify its loan portfolio away from convenience store and gas station loans. This requirement has been imposed on them by the rating agencies if they wish to continue securitizing pools of loans. I would hate for the rating agencies to force me to invest in an area that isn't a core competency of mine (e.g., biotechs) just for the sake of diversification. Seems like an ingredient in the recipe for disaster.
For all you gain on sale fans, you will be interested to know that BLX securitizes the non-guaranteed pieces of the SBA loans and recognizes a non-cash gain on sale at the time of securitization. What are the gain assumptions used by management? Great question but you'll have to ask ALD management - they haven't disclosed the assumptions publicly. This is particularly interesting in light of the fact that companies that use gain on sale are now required to disclose the assumptions used. Apparently ALD doesn't believe it is required to disclose the assumptions given that BLX is not a direct operating subsidiary. Management has disclosed that the residual assets on BLX's books at 9/30/03 were $176 million or 130% of owners' equity.
So what multiple does ALD put on BLX's EBITM? As of 9/30/02, BLX was valued at approximately 7.5x LTM EBITM. As of 9/30/03, BLX was valued at 11.3x EBITM.
How does ALD arrive at an 11.3x EBITM multiple? Nobody is really sure. ALD's BLX peer group has only one company (PMC) that uses gain on sale accounting. The only former peer that did was DVI and it went bankrupt so ALD now excludes it from the peer group. To replace DVI, ALD conveniently added CapitalSource (a richly valued portfolio lender) to the peer group. So basically ALD is taking portfolio lender multiples and applying them to gain on sale earnings. Regardless, BLX's valuation seems aggressive relative to the peers even if BLX was a portfolio lender. As far as private market value, AMRESCO shopped its SBA platform and couldn¡¦t find a bid. ALD ultimately bought loans and residuals (both at a discount to book value) and AMRESCO liquidated the platform.
My opinion is that ALD arrived at an 11.3x multiple by solving for it given they had the EBITM and the desired valuation. Using a 7.5x multiple on LTM EBITM (consistent with last year), BLX equity would be valued at $227 million creating an unrealized LOSS of $34 million. This would create a $121 million DECREASE in unrealized in the current fiscal year wiping out ALD's ENTIRE stated YTD increase in NAV from operations of $113 million, and then some.
You may be asking yourself, 'What's with using EBITM to value a financial company anyway?' Beats me but it is the method the company has consistently used (there actually is a reason they use EBITM but see the Hillman discussion below for the reason). Even if you assumed the appropriate multiple should be 9x (an increase of 1.5x compared to last year), the total unrealized gain would drop to $26 million and create a $61 million decrease in unrealized gain in the current fiscal year wiping out HALF of ALD's stated YTD increase in NAV from operations.
ALD's existing valuation of BLX assumes a 56x P/E multiple on PRE-TAX net income, up from 32x last year. Price to book appears to be 2.2x vs. 2.8x last year. P/B is distorted for two reasons. First, ALD paid 4x tangible book for its initial and most significant investment in BLX. Second, ALD made a significant new investment in equity during the FYE 9/30/03 thus increasing the denominator in the P/B equation. Another way of looking at it is to say that current equity is overly padded with 'fresh' equity. If we strip out the fresh equity from both sides of the equation, P/B as of 9/30/03 is 3.1x. Oh, and don¡¦t forget that residual assets constitute 130% of BLX's equity.
Enough about BLX, you can form your own opinion.
Hillman Companies Valuation
Hillman Companies is a bit more straightforward than BLX. Before I dive into the valuation, let me make an observation about ALD¡¦s valuation methodology. ALD prefers Enterprise Value-type valuations (i.e., multiple of EBITDA or some derivative thereof). ALD's theory is that with them being both a debt and equity investor, it is more accurate to look at their investments on an EV basis because ALD could easily change the mix between debt and equity and that would have an impact on non-EV based valuations. True enough but notice a MAJOR benefit ALD derives from this methodology. As the lender, ALD receives interest income and includes such interest income in its operating income. But notice that the valuation of the investment is not affected by the interest expense as the valuation is derived before interest expense (even for a financial company such as BLX) by using EBITDA. In my opinion, this creates a 'double dip' for ALD's increased operating income for ALD and increased valuations of its investments (which leads to higher unrealized gains which leads to higher NAV and so on). Back to Hillman.
Hillman is a manufacturer of equipment used to make keys and a distributor of key blanks, fasteners, signage and other small hardware components. Hillman files its own 10-K and 10-Qs so information about its financial performance is readily available. The Hillman debt and equity investments (at FMV) represents 6% of ALD's assets, 10% of the private finance portfolio and 10% of NAV. Similar to BLX, it is probably helpful to note the following changes in cost, FMV and unrealized gain (in $millions).
Date Cost FMV Unrealized gain
3/31/02 98 98 0
6/30/02 98 131 33
9/30/02 98 131 33
12/31/02 93 180 87
3/31/03 93 181 88
6/30/03 93 181 88
9/30/03 93 186 93
ALD's basic valuation methodology for Hillman is 7x EBITDAM. ALD arrived at 7x after looking at comps such as Fastenal and determining that the public peer group traded at 10x EBITDAM. In order to be conservative, ALD discounted that 10x EBITDAM multiple by 30% to get a 7x EBITDAM multiple. Further, ALD management has said that they then also apply a 15-17% discount to arrive at the private market discount. In actuality, it appears that ALD uses a 13.5% discount but who is counting.
The problem with Hillman is multi-fold. First, Hillman is basically one giant roll up. The ROIC is pathetic, maybe 8% on a good day. The earnings quality is suspect (cash flow significantly lags net income). In addition, it doesn't seem to be a growth company, short of additional acquisitions.
ALD management has currently valued Hillman at 7x 2003E EBITDAM of $61.5 million. One could debate the multiple used (and I think someone should) but let's stick with the low hanging fruit. For the first nine months of 2003, Hillman generated $41.2 million of EBITDAM. Thus, Hillman needs to generate $20.3 million of EBITDAM in the fourth quarter to meet the $61.5 million estimate. This looks like it might be a little tough to do because EBITDAM would need to increase 67% over last year during the fourth quarter. Last quarter EBITDAM was up 9% YoY. Hillman needs to step on it or ALD needs to come up with an argument for multiple expansion for this roll up.
Hopefully by now I have addressed the inflated assets issue. Again, my theory is that if you think the biggest and most visible investments are booked at inflated values, it's a reasonable bet that the smaller ones are too. These discussions of BLX and Hillman are by no means exhaustive but rather are intended as a big picture overview.
Why short ALD now?
I believe now is an ideal time to short/buy puts on ALD. Here is the short list of catalysts (no pun intended):
(a) ALD cuts the dividend and throws in the towel
(b) ALD is subject to regulatory action, in particular at BLX as has been rumored for some time
(c) ALD's auditors finally clue in as ALD has put itself in a box with respect to the valuation of some high profile investments (e.g., BLX and Hillman)
Here is my assessment of the likelihood various catalysts:
(a) Not likely near-term without some serious prodding. In fact, ALD management has recently told investors that if it didn't have operating income and realized gains to support the dividend in the near-term, it would continue to pay dividends at the same rate out of capital. Scary.
(b) More likely but impossible to time
(c) MUCH more likely and this would happen in the next three months
Given that ALD is on a calendar year, its auditors usually complete the audit in February and ALD reports earnings in late February. It seems to me that if the auditors question the valuations of two of the biggest investments, they may lose faith in the rest of the financial statements. In addition, if the auditors were to require ALD to disclose BLX's gain on sale assumptions, I think it would be enlightening and frightening. Given how material BLX is to ALD's NAV, this disclosure alone could be the catalyst.
So why short ALD now? Because I think you have a possible near-term catalyst and this trade puts the odds in your favor timing wise. Also, if you notice, ALD's stock is near its highs. I think part of that is from tax loss short covering. I am seeing that on a few other stocks too.
From all outside appearances, ALD is in fact running out of juice. Note the massive sequential increase last quarter in ALD¡¦s interest and dividend income. It is very hard to believe this was not from one-time items and management did a poor job of discussing the source of these items on the conference call. Further, as with most investment portfolios, if you need to generate realized gains, you sell what you can but not what you probably should sell. ALD has likely sold all the easy assets that had unrealized gains¡Xcutting its flowers and watering its weeds.
One of the things working against the shorts right now is ALD¡¦s repeated equity issuances in the $25+ level. If you sell stock worth less than $10 to the retail public for $25+, you have increased the true NAV. It¡¦s bad math for the shorts but an unsustainable business model for ALD.
Isn¡¦t ALD expensive to short given that you have to pay the dividend and also have to pay to borrow the stock?
ALD is an extremely expensive short. First, if you can even borrow the stock, you will pay around 15% to stock loan as a negative rebate. Then you have to pay the dividends paid on ALD stock which is about 8.5% right now.
So the solution is to buy the May puts. I have been buying the May 25¡¦s (which trade for $1.25) but the risk/reward on the May 22.50¡¦s (which trade at $.60) may be better. The premium cost for the May 25¡¦s is less than the cost of carrying a short position for three months so you are basically getting a free option as compared with actually shorting the stock (granted the strike prices are $2 different).
This is one of those ideas, I think, that when it works, it will really unravel. ALD has a stated NAV of around $14.50/share. If the asset values are inflated, the true NAV is much lower. If, no when, ALD hits the wall, I believe the shares will trade down to at least the stated NAV and most likely lower. If this trade works this spring as planned, the return on the puts would be $10+.
">
Catalyst
ALD's auditors finally clue in as ALD has put itself in a box with respect to the valuation of some high profile investments (e.g., BLX and Hillman)
ALD is subject to regulatory action, in particular at BLX as has been rumored for some time