The implication of this tortured exercise was that Allied determined BLX was worth nine times EBITM, sixty-five times net income and about five times book value. Sixty-five times net income and five times book value do not pass the laugh test: No one would value a gain-on-sale securitizer that richly. Market values for gain-on-sale securitizers fell so low that once DVI went bankrupt a few months later, we could no longer identify a publicly traded company that generated a large portion of its income through gain-on-sale securitizations.
Table 12.1 Allied’s June 2002 Valuation of BLX
Later in the call, Darius Brawn of Endicott Partners again raised the issue of how much of BLX’s $85 million in revenue came from gain-on-sale. Sweeney indicated that she didn’t know off the top of her head, but after Brawn pressed her, Sweeney said that the cash portion of BLX’s revenues was around 50 percent to 70 percent. This meant that the 30 percent to 50 percent of $85 million in revenues, or $25 million to $42 million, was non-cash. Since gain-on-sale revenue generally flows through the income statement without marginal expense, this meant the majority of the $43 million of EBITM was non-cash.
For the first time, Allied provided some summary financial information for BLX in its 10-Q filed weeks later. Since the last available disclosure about BLX at its formation in December 2000, BLX’s debt increased $65 million to $183 million and an intangible asset created by gain-on-sale accounting called “residual interests” roughly doubled to $106 million. As the residual interests were almost four times the tangible equity, you could call the valuation whimsical, or imaginative, or fanciful—but not supportable.
Putting it all together, BLX was a problem in three ways. First, BLX made loans improperly. Second, BLX used aggressive accounting techniques to inflate its results. Finally, Allied’s valuation of BLX had no reasonable basis, even if BLX’s business and accounting results had been genuine. Frankly, as we were learning more about BLX, we were coming to believe Allied’s investment could really be worthless.
Allied sought to convince investors it wasn’t a fraud by giving evidence that it didn’t behave like other frauds. Allied paid distributions, proving the cash wasn’t missing. Insiders purchased shares, signaling the market that nothing was amiss. It had consistent accounting, validating its four decades of success.
It was like the old Richard Pryor joke: “What do you do if your wife walks into your bedroom and finds you in bed with another woman? Deny! Deny! Deny! Who are you going to believe? Me or your lyin’ eyes?” So, too, with Allied.
Despite the large changes in Allied’s accounting practices, to maintain the confidence of inattentive or unsophisticated investors, management repeatedly advanced the illusion that Allied’s accounting was consistent. During the second quarter of 2002 earnings conference call, Walton led off his discussion of valuation, “as a result of our consistent process of determining the fair-value of our portfolio in good faith . . .” Moments later, Sweeney echoed, “As we stated in our public filings, we used a consistent methodology to value our portfolio companies in keeping with the guidance provided by the SEC and industry practices.” Before turning to the Q&A, Walton reiterated, “Let me emphatically state that we will continue to apply a consistent and prudent valuation methodology that is in accordance with all regulatory guidelines as we have always done.”
The evidence that Allied changed its accounting is overwhelming. There were remarkable sudden changes to non-performing assets, PIK and fee income, and write-downs and write-ups. A few weeks later, Allied dramatically changed the narrative description of its accounting in its new 10-Q.
Allied eliminated the questionable language that echoed the white paper that first appeared in the 2001 annual report. The company replaced it with new prose describing its use and interpretation of the current-sale test based on enterprise value. The June 2002 10-Q declared for the first time, “The fair-value of our investment is based upon the enterprise value at which the portfolio company could be sold in an orderly disposition over a reasonable period of time between willing parties other than in a forced or liquidation sale. The liquidity event whereby we exit a private finance investment is generally the sale, the recapitalization or, in some cases, the initial public offering of the portfolio company.”
Over time, the effects of the new accounting persisted. Fees, interest, and PIK income stayed lower. The number of write-ups and write-downs recognized each quarter remained more pronounced. Earnings that had been growing steadily and predictably became volatile and unpredictable. Net investment income per share, which grew in a straight line for several years through the first quarter where they reached fifty-three cents per share, reversed their steady upward trend and became more volatile and stopped growing. They have yet to return to anywhere near that first quarter of 2002 level (see Figure 12.1).
Figure 12.1 Quarterly Net Investment Income
Allied not only changed how it evaluated its portfolio, but also changed how it wanted investors to evaluate the company. In the past, Walton directed investors to focus on recurring net investment income (also called operating income) and observed that capital gains were nice but unpredictable. With the new accounting, Allied’s net investment income no longer approached the quarterly distribution. Instead, Allied refocused investors on taxable income, which consists of net investment income plus realized capital gains. To that end, Walton announced that the company would no longer give earnings guidance and from then on would only provide “dividend” guidance for its tax distribution.
“The dividend, as you know, is based on taxable earnings,” he said. “We find that the timing differences between tax and GAAP earnings result in our GAAP earnings being less meaningful to shareholders. . . . The reality for Allied Capital and its shareholders is taxable income supports the dividend, and although GAAP earnings are useful as a predictor of