I have watched AutoChina (AUTC) and The Forensic Factor's (TFF) public dialogue with interest (the links to the articles and responses are at the bottom of this article). TFF unearthed some interesting and poignant concerns on accounting, capital, dilution, related party transactions, and management background. However, management's response was quick and thoughtful. TFF must have really struck a cord because in a shareholder friendly move, AutoChina's Chairman and CEO agreed to revise the earnout hurdles that could limit dilution over the next few years.
I applaud the move, as I believe it was required for investors to revisit the shares. However, I thought it was disingenuous to include 2010 revised hurdles a month and a half after 2010 ended, especially since the hurdle was set well below analyst estimates. Clearly, management should be commended for its actions, but that does not make AutoChina a good investment.
According to chat rooms and market sources (I cannot confirm), a large online retail brokerage firm lent over 500,000 share of AutoChina stock to short sellers. When the brokerage clients sold their shares recently, the firm required the shares back. Over the last week, short sellers have been forced to find a new borrow (none available at Schwab or Morgan Stanley) or purchase the shares in the open market. Since February 22, 2011, when the stock was near its 52 week low, the shares have surged over 50%. This is one of the most violent "short squeezes" I can recall. Based on recent volumes, I estimate that this squeeze should be nearing an end (if the share amounts are correct). If this is the case, this artificial demand for shares could disappear and the stock could be in for a painful correction.
I believe management avoided a few very key points in their responses, and I highlight some of them in this article. It seems from management's responses that they will complete the audit soon. At the very least, I think management should address these concerns on their upcoming earnings call. What is clear from my research is that AutoChina's sell side analysts have mis-modeled the company's financials and have likely overstated future earnings. What is also clear is that AutoChina needs access to large amounts of debt and equity capital. I believe management will have to address the following 12 questions before U.S. investors should be willing to provide management with that capital.
1) What are AutoChina's specific capital needs? And what are the plans to issue equity that could further dilute shareholders? I believe this is the most important question that TFF brought up and was more or less ignored by AutoChina's management.
AutoChina operates an extremely capital intensive business (it admits as much). It appears that analysts covering the company have little concept or experience in modeling financial services companies. It appears that their estimates are wildly optimistic, providing the company full benefit of large origination growth while not accounting for how the company will fund that growth.
An example of this is in the recent initiation of coverage from Daiwa in January 2011. The analyst assumes 20,625 originations in 2011 and 28,750 in 2012 to produce 62% and 39% revenue growth, respectively. At the company's (and TFF's) cost per vehicle of $38,000 and 20% down, this equates to $627 million of capital needs in 2011 and nearly $874 million of capital needs in 2012. If AutoChina has perfect credit, it estimates that it will collect roughly 40% of this amount from the existing portfolio in 2011, leaving a $336 million funding gap this year. The Daiwa analyst is projecting a $95 million increase in debt in 2011 and $180 million in 2012. The analyst does not explain where the additional $241 million will come from and it does not appear to be reflected in the model.
Surprisingly, the analyst has only modeled a $5 million increase in interest expense in 2011 ($19 million total) and only a $3 million increase in 2012. AutoChina was already on a $19 million annualized interest expense run-rate in the third quarter 2010. I have no explanation for Daiwa's analysis. A relevant analogy would be if an Apple (AAPL) analyst had new iPad sales at nearly 100% incremental margins (and ignoring costs of sales). Adding $5 million of interest expense to AutoChina's securitization trust at 9% implies $55 million of additional debt (or only 16% of needed capital). The Daiwa model does not include an equity raise to fund originations, as his 23.5 million shares appear to include new shares issued under the 2010 incentive plan. Daiwa's model therefore appears to assume the company can fund $336 million of originations for free.
If I assume the company could fund this $336 million funding gap at 8% debt, this would equate to $27 million of additional interest expense, or $0.86 per share. This change would take Daiwa's estimate from $2.51 per share to $1.65 per share (so trading at nearly 20 times earnings). What is even more surprising is that Daiwa does not have the high estimate on the street. There is another analyst estimate of $3.00 per share for 2011. What is clear is that these estimates are way too high and AutoChina is set up for a spectacular earnings miss.
This debt analysis above is more for show since the company cannot raise $336 million in additional debt (they would be too overleveraged). In addition, availability of capital is a concern in China right now. China's central bank is tightening liquidity to slow down an overheating economy. Chinese officials recently raised the reserve ratio on banks to an all-time high of 19.5%. Chinese banks are being discouraged from providing loans to private businesses in an effort to control inflation.
Evidence of this was on display last week when a highly profitable, but highly capital intensive business, LDK Solar (LDK), needed to raise debt. Historically, LDK was able to access debt capital from Chinese based banks at around 5%. Last week, they issued RMB 1.2 billion ($185 million) of two year debt at a whopping 10% interest rate. Given the large rate difference, it leads me to believe that cheaper bank debt was no longer available to LDK. Are they available to AutoChina?
The only way for AutoChina to grow may be to issue equity capital (which is not reflected in analysts' models). If not, they will be forced to stop or reduce originations (and earnings will drop). The book value of the company is only $10.96 per share ($9.13 adjusted for the 2010 earn-out), so a stock issuance at a large discount to current market will still be accretive to book value. Since analysts have mis-modeled the company, equity issuance will very, very dilutive to the current earnings outlook.
I estimate that the majority shareholder, CEO Yong Hui Li, will have a cost basis of $5.69 per share after he receives his 2010 earnout (8.6 million issued at $10 per share from the original SPAC plus 6.5 million earnout shares at zero cost). As a result of his low cost basis, he may be more willing to issue shares to a public at lower prices.
2) If AutoChina needs to fund over $600 million of originations, how much of that will come from a) new equity issuance b) interest and principal collections c) new debt?
3) With significant leverage already on the balance sheet, how high can leverage go (on a debt to equity basis)?
4) What would the 2010 income statement look like if the company expensed related party debt at fair value? Why does APB Opinion No. 21 not apply to AutoChina's financial statements? This is an area not explored by TFF, but appears to be another accounting treatment that could be a cause for restatement. Under APB Opinion No. 21, debt with below market interest rates must be discounted. The purpose of discounting is to adjust the note to an equivalent note having the market rate of interest. The discount is then amortized through the income statement as an interest expense.
AutoChina's management does not hide the fact that they receive zero interest debt from companies owned by the CEO and below market rates from other companies partially owned by management. All of the related party debt appears to be below market rate of interest (of their bank debt). APB No. 21 does not apply for payables due within one year under the "normal course of business". But AutoChina's debt is not a payable related to something like an invoice for computer supplies ("normal course receivables"), it is to fund its operations. As a result, AutoChina's income statement is significantly understating its interest expense.
My professional colleague, Eric Friedman at Cornell University, labeled this business practice "propping" in his 2003 paper (the link is here (.pdf)). Another university paper described propping as when "one entity undertakes transactions that are otherwise not in its own economic interest, but serve only to maintain the financial health of another entity" (that link here (.pdf)).
Friedman found that
issuing debt can credibly commit an entrepreneur to prop, even though creditors can never take possession of any underlying collateral. This helps to explain both why emerging markets with weak institutions sometimes grow rapidly and why they are subject to frequent economic and financial crises.
I also found the following article interesting: "The Role of Related Party Transactions in Fraudulent Financial Reporting". It found that "loans to related parties were the most frequent type of related party transaction" that drove SEC enforcement actions involving fraud.
5) How much does it cost to operate a store / lending office and how many employees does the company have? How many employees do you have per store? This is another area that TFF did not explore but appears questionable. In 2009, AutoChina disclosed 1,206 employees and 157 "stores." On December 31, 2010, AutoChina had 300 stores and I estimate over 2,000 employees based on lower employees per store than 2009.
This represents $13,000 to $15,000 of annualized SG&A per employee through nine months 2010. The level of compensation per employee would be much less after deducting rent, stock comp, depreciation and amortization, equipment, insurance etc from the SG&A figure and is below levels at other Chinese companies. This appears to be another questionable data point after analyzing the company's financial statements.
6) Please explain why, as a finance company (and not a manufacturer or captive leasing company), you utilize sales type lease accounting? This issue continues to remain unclear.
7) What would the company's third quarter and year-to-date revenue be if it did not use sales type lease accounting? In TFF's original report, they provided an example where TFF estimated AutoChina overstated revenue by 92.5% over the first three months of new lease. On an overall basis, I estimate that third quarter revenue would be roughly 55% lower without "sales type lease" accounting. Is this correct?
8) What would the company's 2010 earnings be if it did not use sales type lease accounting?
9) Please explain how classifying delinquent accounts as accounts receivable conforms to GAAP? Also, please provide disclosure on how much of your $1.1 million allowance for doubtful accounts relates to specific reserves and general (future) reserves.
10) How is EBITDA calculated for the earnout? EBITDA is a proxy for cash flow, but AutoChina's operating cash flows have been very negative over the past six quarters. I have never seen EBITDA used as a performance metric for the finance company, especially one that uses leverage. With EBITDA, a finance company's primary expense, interest on it funding sources is excluded. For example, Citigroup (C) has $11 billion of income in 2010. By adding back $25 billion cost of debt, its EBITDA would be greater than $38 billion (or a useless figure for a financial company). Please explain why EBITDA is relevant?
11) Do AutoChina's shareholders own the operating companies? This point remained unclear in regard to the VIEs. If not, how can this be changed?
12) What would the 2010 income statement look like if the company expensed the earnout?
AutoChina's recent stock performance and valuation suggests the company is healthy and without significant questions. In my analysis of the company, it appears that significant issues remain and management must address these concerns. I believe the recent stock performance relates to a short squeeze that is likely coming to a end. As a result, I see large downside for AutoChina shareholders in the near-term.
Links and timeline:
February 1, 2011: "The Most Preposterous Chinese Reverse Merger Yet" (.pdf)
Feburary 3, 2011: Company response #1(.pdf)
February 7, 2011: "Imagine if this was a U.S. company?"
February 8, 2011: Company response #2
Disclosure: I have no positions in any stocks mentioned, but may initiate a short position in AUTC over the next 72 hours.