LOSS CALCULATION IN RE SENTENCING

Before the Court is an issue raised during the sentencing of Defendant Osamah S. Bakhit on eighteen counts of securities fraud, bank fraud and related violations in connection with the accounting practices of his company, Aviation Distributers, Inc. (ADI). The Government and the United States Probation Officer have put forth two separate theories for calculating the loss attributable to Defendant’s fraudulent activity and Defendant counters with his own theory and expert testimony.

The base offense level for fraud is fixed at 6, but can increase by as much as three times, to 24, when loss is taken into account. See generally U.S.S.G. § 2F1.1 (Nov.1997). Under this offense level range, Defendant’s sentence could vary by as much as five years. Yet, the basis for this vast disparity in sentence is governed only by the United States Sentencing Commission Guideline’s (Guidelines) vague standards of loss. This Court’s order illustrates the myriad approaches to, and the numerous obstacles that come with, finding the loss attributable to harm at the sentencing stage.

After reviewing the papers submitted in this matter, and for the reasons set forth below, the Court finds the loss to be $2,884,200, resulting in a 13 level upward adjustment.

I. BACKGROUND

Defendant is the founder and former CEO, President, and Chairman of the Board of Directors of ADI, headquartered in Orange County, California. ADI is a supplier, distributor and broker of used and new commercial aircraft parts and supplies and its customers include commercial airlines around the world.

Beginning in 1994, and continuing into 1997, Defendant was involved in fraudulent billing practices at ADI. The fraudulent billing included bogus invoices, rebilling, and fraudulent inventory exchanges which resulted in falsely inflated revenue. In order to finance its operations, ADI maintained a borrowing relationship with Far East National Bank (FENB) and the fraudulent billing allowed ADI to obtain cash from FENB on its line of credit earlier than it would otherwise have been available. Additionally, the fraudulent statements were included in ADI’s 1994, 1995 and 1996 financial statements. As a result of these practices, ADI appeared to have attained and maintained a healthier financial condition than in actuality.

In 1996, Defendant and ADI began planning to take the company public. ADI retained Cruttendon-Roth (CR) to underwrite the initial public offering (IPO) and in February 1997, ADI filed a registration statement with the Securities Exchange Commission (SEC). The registration statement sought approval for the sale of 1,380,000 shares of common stock to the public and it included the financial statements for 1994, 1995, and 1996 that reflected Defendant’s fraudulent billing practices. Relying upon these statements, the SEC authorized the IPO.

On March 3, 1997, ADI conducted its IPO, selling 1,380,000 shares of common stock to the public at a price of $5.00 per share. The total proceeds of the IPO were approximately $6,900,000 and ADI used a substantial portion of the money to repay amounts outstanding with FENB. As a result of the IPO, ADI’s common stock began public trading on the National Association of Securities Dealers Automated Quotation (NASDAQ). Over the next several months, the price of ADI’s shares increased steadily, reaching a high of approximately $12.00 per share in August 1997. During this time period, Defendant did not sell a single share of ADI stock.

In early August 1997, ADI’s auditor, Arthur Andersen, discovered a falsified invoice in ADI’s books and records. On August 29, 1997, Arthur Andersen resigned as ADI’s independent auditors and withdrew its audit reports for ADI’s 1994, 1995 and 1996 financial statements. In its resignation letter, Arthur Andersen stated that it believed that ADI had prepared false sales invoices and documents, providing them to FENB for financing as well as to Arthur Andersen. Arthur Andersen further noted that “we believe that the foregoing matters have occurred with the knowledge and involvement of the highest levels of management in the company.” On August 30, 1997, trading in the company’s stock was halted and, in October 1997, NASDAQ delisted ADI. In November 1997, ADI retained Gram Grant Thornton LLP as its new auditor and Grant Thornton LLP required Defendant to resign as Chairman, CEO and President. Despite the resignation, Defendant continued, and continues to date, to work for ADI as a consultant, salesman, and rainmaker.

Trading resumed in January 1998, with ADI’s stock selling at approximately $6.00 per share after a single, initial drop to $3.00 per share. On April 20, 1998, ADI restated its 1996 financial statements. ADI’s 1996 net income was reduced from a profit of $315,000 (net profit of $0.18 per share) to a loss of $1,566,000 (net loss of $0.88 per share). Aso in April, ADI announced its settlement of a securities fraud class action lawsuit. The settlement consisted of $740,000 in cash and 210,000 shares of stock. Defendant’s contributed stock to the settlement fund and this contribution was his first disbursement of any of his ADI holdings. Trading remained steady at approximately $5.00-6.00 per share through the earnings restatement and through ADI’s announced settlement with class members, until mid-May 1998. The next marked drop in stock price occurred about May 17, 1998, when ADI announced its first quarter 1998 earning results.

On February 11, 2002, Defendant pleaded guilty to Counts 1 through 13, 16, and 18 through 21 of the Indictment, including multiple counts of fraud and false statements in relation to securities and accounting practices. On July 15, 2002, the United States Probation Officer disclosed the Presentence Report (PSR), recommending a fourteen level enhancement pursuant to Section 2Fl.l(b)(l)(0) based upon a loss of $6,900,00o. The Government concurs but Defendant objects, claiming that the PSR’s calculation of loss overstates the actual loss attributable to the fraud. Defendant argues that there is actually no loss and that there should be no enhancement. A fourteen level enhancement could add as much as three years to Defendant’s sentence.

II. DISCUSSION

A. Section 2Fl.l(b)(l)

Pursuant to Section 2F1.1(b)(1) of the Guidelines, the base offense level for fraud is increased according to the amount of loss attributable to the fraud. As a potentially disproportionate enhancement is involved, the burden is on the government to prove by clear and convincing evidence the amount of “harm that resulted from the acts or omissions” of the defendants. See United States v. Munoz, 233 F.3d 1117 (9th Cir.2000); see also United States v. Hicks, 217 F.3d 1038, 1049 (9th Cir.2000); U.S.S.G. § 2B1.1. The Guideline’s comments emphasize that the court need not determine the loss with precision. Instead, “the court need only make a reasonable estimate of the loss, given the available information.” U.S.S.G. § 2F1.1 cmt. 8. In addition, the court may take either the intended loss or the actual loss, whichever is greater. See U.S.S.G. § 2F1.1 cmt. 7 (“[I]f an intended loss that the defendant was attempting to inflict can be determined, this figure will be used if it is greater than the actual loss.”); see also United States v. Joetzki, 952 F.2d 1090, 1096 (9th Cir.1991). The Ninth Circuit, specifically, instructs district courts “to take a realistic, economic approach to determine what losses the defendant truly caused or intended to cause.” E.g., United States v. W. Coast Aluminum Heat Treating Co., 265 F.3d 986, 991 (9th Cir.2001). Simply put, the district court is not obligated to find “the perfect theoretical or statistical fit.” Id. The Guidelines do, however, contemplate a causation requirement and the government must therefore prove that the defendant’s conduct was the proximate cause of the loss. Hicks, 217 F.3d at 1048-49 .

B. IPO Theory

The PSR assumes, and the Government contends, that the total proceeds from ADI’s IPO are an accurate representation of the amount of loss attributable to Defendant’s fraud. The Government argues that if the fraud had been disclosed prior to the IPO, the IPO would not have occurred and, thus, the full proceeds are the investors’ “loss.” The PSR more fully explains that if the SEC had known the statements were fraudulent, it would have invalidated the IPO and, therefore, the amount of loss is $6,900,000, the total proceeds of the IPO.

The Court finds several problems with this theory. Most importantly, the objective evidence of ADI’s stock price indicates that the initial shareholders’ investment was not completely worthless. To attribute the total proceeds of the IPO to Defendant’s conduct, would require that the Court assume that this is the case. Yet, after the fraud was fully disclosed and absorbed by the market, ADI stock continued to trade at around $5.00 per share. This is not a situation where the shareholders received worthless stock, but rather received stock worth less than they anticipated. In fact, Section 2F1.1 explicitly contemplates its application to an offense involving securities fraud and among the factors offered for consideration, is whether the fraud merely misrepresents the value of an item that does have some value. See generally U.S.S.G. § 2F1.1 cmt. 7, 7(a). “Where, for example, a defendant fraudulently represents that stock is worth $40,000 and the stock is worth only $10,000, the loss is the amount by which the stock was overvalued (i.e. $30,000).” U.S.S.G. § 2F1.1 cmt. 7(a). In accordance with the Guidelines, the actual “loss” in this case should be the amount by which the stock was overvalued, not its entire value (i.e. the full proceeds from the IPO).

Second, the IPO theory does not take into account shareholders who did not buy on March 3, 1997, but rather purchased stock at a later time in 1997. As noted above, the price of ADI stock steadily increased from March 3, 1997 until September 2,1997, the date trading halted. Many of these shareholders purchased ADI stock at a price well above $5.00 per share and their loss is likewise directly attributable to the fraud. The shareholders trading in 1997 relied not only upon the fraudulent statements related to the IPO, but later fraudulent statements as well. Defendant was responsible for fraudulent billing, accounting and reporting that occurred as late as August 1997. This entire time period was prior to disclosure of the fraud and, thus, the trading was impacted by the materially misleading statements. Although the IPO theory may result in a higher amount of loss than a calculation that takes into account the variable buying and selling, it is not a reasonably accurate reflection of the victims’ actual loss.

Alternatively, the Government has not made a persuasive argument that the total proceeds of the IPO were Defendant’s intended loss. Unlike a forged check that is never cashed but has the intended loss amount clearly written on the front, e.g., U.S.S.G. § 2F1.1 cmt. 7, it is speculative to assume that the harm intended was the amount of the IPO. Defendant did not pocket the proceeds or force a buyout of his own stock with the proceeds, either of which would indicate a purpose to defraud shareholders of the total IPO amount. Rather, Defendant paid off amounts due under the FENB lines of credit and, presumably, used the money in the continued operation of ADI, which is in business today. Again, the share price after disclosure of the fraud suggests that, at most, Defendant intended to inflate the stock price, not create a valuable stock out of a worthless commodity. The total proceeds include investment money that is attributable to factors and conduct other than the fraudulent statements, namely legitimate business actions and business entity, that should not be included as intended loss.

In sum, the Government has not shown by a preponderance of the evidence either that the actual loss was $6,900,000 or that Defendant intended the loss to be $6,900,000.

C. Market Capitalization Theory

The Government’s alternate theory of loss is based on “the reduction in market capitalization after the fraud was uncovered and publicized.” The amount of loss is determined by comparing the difference in ADI’s stock price, multiplied by the number of outstanding shares, immediately before the disclosure of the fraud with the stock price just after disclosure of the fraud. According to the Government, ADI’s market value just before disclosure was approximately $16,700,000 ($12,125 per share multiplied by 1,380,000 shares) and the value after disclosure was approximately $8,300,000 ($6.00 per share multiplied by 1,380,000 shares) for a difference of $8,400,000. As support for the market capitalization theory, the Government points to two recent Court of Appeals decisions, United States v. Moskowitz, 215 F.3d 265 (2d Cir.2000), and United States v. Hedges, 175 F.3d 1312 (11th Cir.1999). This Court, however, finds these cases to be unpersuasive and further finds that the Government’s market capitalization formula does not accurately reflect the actual loss in this case.

The Government argues that the Hedges court calculated the loss based upon the loss in share value after the fraud was revealed. In Hedges , however, the defendants’ misrepresentations were so egregious that when the fraudulent conduct came to light, the stock immediately became worthless. 175 F.3d at 1313 . The court noted that the defendant’s conduct created a market for worthless stock and, thus, the loss was the total amount the public was induced to invest. Id. at 1315 . Unlike the instant ease, the Hedges court was not faced with stock that had continued value after the fraud’s disclosure. See, e.g., United States v. Snyder, 291 F.3d 1291, 1296 (11th Cir.2002) (distinguishing Hedges on the basis that the stock in the case before the court was not completely worthless after the conspiracy was disclosed). Further, the Government oversimplifies the Hedges calculation. There, the court considered the average trading price at which the outstanding shares were bought and did not, as the Government suggests, refer solely to the last trading price prior to the fraud’s disclosure. Id. at 1314 n. 6. The Hedges formulation, in fact, more accurately reflects the actual loss than the Government’s calculation, by taking into account the variable prices at which the stock was purchased during the life of the fraud.

Moskowitz is similarly unpersuasive. It is unclear from the Second Circuit decision exactly how the district court reached the loss amount in the underlying case. The Circuit court simply found that there was “evidence that would have permitted a reasonable fact finder to determine actual losses to be higher than the range of $5 million to $10 million that the district court found applicable to [the defendant].” Moskowitz, 215 F.3d at 272 . The court additionally noted that “the government estimated the loss at from $7.1 million to $18.3 million based on the decline in [the company’s] share price upon the revelation of the fraud” and that an expert in a related class action determined the loss to be $30 million. Id. There is, however, no explanation of how the government or the expert arrived at those estimates and no suggestion of the district court’s final calculation. Moskowitz therefore only tenuously supports the Government’s theory and it certainly does not address this Court’s concern with using the Government’s market capitalization analysis.

The Court’s primary concern is similar to one raised in regard to the IPO calculation. The Government’s formulation does not take into account the varying prices at which ADI shareholders purchased ADI stock. Although the market capitalization theory purports to reflect the amount at which the company was overvalued, it overstates the actual amount of loss to the victims. The majority of shareholders purchased ADI stock at a price lower than $12.12 per share, the trading price just prior to revelation of the fraud. Thus, multiplying $12.12 by the total amount of outstanding shares and then subtracting the value just after disclosure, overstates the actual loss by a significant amount.

Additionally, by using only the trading price just prior to the fraud, there is an increased risk that outside factors affecting share price will be given undue weight. Publicly traded stock is inherently volatile and a calculation that focuses on a single day of trading may be more representative of outside events than either the overvaluation due to the fraud or the true value of the company after disclosure of the fraud. As stated above, the loss attributable to the fraud-and used to enhance the offense level pursuant to Section 2F1.1-must be proximately caused by Defendant's conduct. E.g., Hicks, 217 F.3d at 1048-49 . The calculation of loss must therefore account for significant intervening causes. Id. A calculation that encompasses a broader range of trading and share prices would lessen the impact of outside factors and, likely, result in a more accurate estimate of actual loss than the Government's market capitalization theory.

D. Defendant's Expert Calculation

In addition to opposing the Government's calculations, Defendant has put forth his own theory for loss calculation. Defendant relies upon the calculation of a civil securities fraud expert that purports to match trading purchases with sales to arrive at the actual loss. Dr. Cornell asserts that his method effectively offsets the gains realized by some shareholders against the losses suffered by others and, assuming that the true value of ADI stock after the fraud disclosure is $6.00 per share, arrives at $888,602 as the net loss suffered by ADI shareholders. (Cornell Dccl. ¶114-7.) Dr. Cornell then multiplies the net loss by 17.47%, as Defendant contends that only 17.47% of ADI's total restatements between 1995 and 1997 can be attributed to fraudulent financial transactions (as opposed to sloppy bookkeeping and other mistakes). (Id. Ii 7.) Dr. Cornell concludes that the loss attributable to Defendant's actions is only $155,239. (Id.)

There are two significant problems with Defendant's position. First, by offsetting one shareholder's gain against another's loss, the loss calculation rewards Defendant for inflating the price such that initial investors were able to sell their stock for more than its actual value. A similar approach was rejected by the Eleventh Circuit in United States v. Orton, 73 F.3d 331 (11th Cir.1996). There the defendant argued that the net amount lost by all victims in a Ponzi scheme was only $283,352.01 and, accordingly, that should be the amount of loss under Section 2F1.1. Orton, 73 F.3d at 333-34 . The total amount lost by victims who suffered individual net losses, however, was $391,540.01. Id. at 333 . The court refused to offset the gains of those victims that received more money back than originally invested, against those that received less. Id. at 334 . The reason for using the total amount lost, the court noted, is that it "does not allow the defendant to fully benefit from payments made to others." Id.

Although OrIon involved a Ponzi scheme, where giving some early victims a high return on their investment is essential to the commission of the fraud, it is still instructive. The initial increase in ADI's stock price was premised upon fraudulent conduct and it similarly enticed later victims to buy ADI stock. These shareholders ultimately lost much more than the early shareholders who purchased at the IPO price. The policy against rewarding a defendant for continued concealment of the fraud that allows some victims to prosper, cautions against using Dr. Cornell’s calculation.

Second is a prudential and political concern with relying too heavily on expert testimony. Although it may be preferable for the district court to have the benefit of dueling experts and an extensive tutorial to determine the actual loss with exactitude, that is simply not practical in the vast majority of criminal fraud cases. Most defendants do not have the resources to hire an independent expert and the government has similar financial constraints. Further, when expert testimony involves complicated algorithms, a court may place too much emphasis on a single expert’s opinion if it is not also presented with an opposing or independent expert to explain and counter the conclusions. Allowing and encouraging expert calculation of loss brings sentencing proceedings one step closer to being a trial in all but name. Yet, Defendant’s guilty plea, with the concurrent acceptance of responsibility, should minimize the resources expended in proving the offense and provide Defendant some benefit for his cooperation. The Guidelines allow the loss calculation to play a pivotal role in sentencing and have already placed a substantial burden on the parties and the Court to determine significant, fact-intensive issues after the entry of a guilty plea. Relying upon expert testimony would only compound the problem and increase the burden.

Finally, the Guidelines stress that the court need only make a reasonable estimate based upon the available information. U.S.S.G. § 2F1.1 cmt. 8. Turning to an expert, who makes substantial assumptions and applies a complex formula, would undermine this principle.

E. The Court’s Calculation-Average Victim Loss

The above critique of the Government’s proffered loss calculations and the Court’s rejection of Defendant’s expert calculation, leads the Court to adopt an entirely separate loss analysis based upon the available information. Following the dictates of the Guidelines and a suggestion by the Eleventh Circuit, the Court attempts with this calculation to reasonably estimate the actual loss.

According to the Guidelines, the court’s estimate “may be based on the approximate number of victims and an estimate of the average loss to each victim, or on more general factors....” U.S.S.G. § 2F1.1 cmt. 8. In a securities fraud case involving overvalued stock, the Eleventh Circuit implicitly followed this instruction. See generally United States v. Snyder, 291 F.3d 1291 (11th Cir.2002). In Snyder , the defendants released a false press release touting the efficacy of BioCryst’s new cancer drug. Snyder, 291 F.3d at 1293, 1296 . BioCryst stock jumped from around $6.25 to $8.25 per share and continued to climb in exceptionally heavy trading for about two weeks. Id. at 1296 n. 6. During this period, BioC-ryst stock averaged $9.63 per share. Id. At the end of the two week period, the false statements were revealed and the stock dropped by as much as $6.00 per share, averaging $8.13 per share for the three days following the announcement. Id.

Like the instant case, the BioCryst stock was not worthless after fraud disclosure, but simply fell in value, indicating that the fraud artificially inflated the stock’s trading price. The Circuit therefore suggested that the district court take the difference between the average price during the fraud ($9.63) and after disclosure of the fraud ($8.13), for an average loss per victim ($9.63 - $8.13 = $1.50). Id. The average loss per victim is then multiplied by the total number of victims. See id. Noting that this was a particularly focused approach, due to the relatively short duration of the fraud, the court more generally noted that “subtracting the stock price after the fraud from the average stock price over the life of the fraud — may be appropriate in many cases involving securities fraud.” See id. at 1296 .

Given the objectivity of the approach and its adherence to the dictates of the Guidelines, the Court finds the Snyder court’s calculation to be a reasonable and appropriate approach in a securities fraud case involving overvalued stock. In the instant case, the calculation is as follows.

1. Average Selling Price During Fraud

The first step is determining the average selling price during the life of the fraud. Here, the life of the fraud is the period from March 3,1997 to September 2, 1997. March 3, 1997 was the first day of trading and from that initial IPO, ADI’s public records and other financial statements were tainted by Defendant’s fraudulent practices. September 2, 1997 is the last date of trading prior to the market impact of Arthur Andersen’s resignation, with its accompanying disclosure of fraudulent statements and the resulting halt in trading. Based upon the closing price on each day of trading, the average trading price during the life of the fraud was $7.64 per share.

2. Average Selling Price After Fraud

The next step is the average selling price after disclosure of the fraud, which should reflect the true value of ADI. After the fraud disclosure, trading was halted for approximately five months and did not resume until January 20, 1998. Trading on January 20, 1998 closed at $3.00 per share but over the next several weeks, from January to mid-May, the stock traded steadily at approximately $5.00-6.00 per share. The trading on January 20, 1998, appears to be an anomaly, an extreme reaction to the announcement of fraud. Because the stock rebounded and then traded steadily for several weeks after January 20, 1998, the entire period from the resumption of trading until mid-May is included as the period after disclosure of the fraud.

The next significant drop in stock price occurred on May 18, 1998, in response to ADI’s first quarter 1998 earning results, announced on May 17, 1998. The reduction in value after the May 17, 1998 announcement should not be included in the loss analysis as attributable to Defendant’s conduct, as the earnings announcement is a significant intervening cause.

The average selling price after disclosure encompasses the period from January 20, 1998 until May 13, 1998 (the last day of trading prior to the earnings announcement). Using the closing price for each day, the average trading price during that period was $5.55 per share.

3. Total Loss

The third step is calculating the difference between the average selling price during the fraud ($7.64) and the average selling price after the fraud ($5.55), which is $2.09 per share ($7.64 - $5.55 = $2.09). This represents the amount by which the stock was artificially inflated and the amount lost per share. As all of the shares were purchased during the life of the fraud, each one was subject to this overvaluation and the subsequent loss. The difference ($2.09) is thus multiplied by the total number of outstanding shares (1,380,000) for a total loss of $2,884,200. Pursuant to Section 2Fl.l(b)(l)(N), a 13 level enhancement is applied to the base offense level of 6 because the loss is more than $2,500,000 but less than $5,000,000.

J. Summary

As stated above, the district court need not find a perfect theoretical model to calculate loss, only a realistic, economic approach. West Coast Aluminum, 265 F.3d at 991 . Although this Court’s approach — using average trading prices— sacrifices some precision, it alleviates many of the problems inherent in the Government’s and Defendant’s loss calculations. The calculation takes into account shareholders who purchased at a price higher than the $5.00 per share IPO price and those that purchased at less than $12.12 per share. It does not reward Defendant for some shareholder’s gains but it does reflect that the stock retained some value after the disclosure. The calculation encompasses the entire period of the fraud and allows for an extended or belated response to the fraud disclosure. Also, the Court can, without the aid of expert testimony or an extensive factual debate, calculate the loss based upon readily available information. The calculation is based upon objective trading data, easily obtained, that minimizes the speculation found in the other proffered calculations.

As a final note, this loss amount parallels the settlement amount in a shareholder class action involving much of the same conduct. The settlement was for $740,000 in cash plus 210,000 shares of stock (trading at approximately $5.50 per share at the time of settlement) for an approximate total of $1,895,000. Considering the discount that a settlement award includes— reflecting the risk of litigation, the expense of litigation, and the parties’ compromise— the Court’s loss of $2,884,200 is proportionate to the settlement.

III. CONCLUSION

The Court finds the total loss attributable to Defendant’s fraudulent conduct to be $2,884,200, resulting in a 13 level increase pursuant to Section 2Fl.l(b)(l)(N).

IT IS SO ORDERED.

. The November 1, 2001 edition of the Guidelines Manual provides for a higher adjustment for loss and thus creates an ex post facto problem. Therefore, the November 1, 1997 version of the Guidelines Manual, which was in effect at the time of the offense, is used to calculate the offense level. All citations herein are to the November 1, 1997 edition of the Guidelines. . Although ADI's 1996 earnings were not restated until April 20, 1998, by that point the market had been aware of the fraudulent statements for a significant period of time and it appears that ADI stock was no longer overvalued based upon the false statements. Indeed, on April 20, 1998, the market did not noticeably react to the announcement, decreasing a mere $.26 per share. The Court, therefore, finds September 2, 1997, the last date of trading prior to the halt in trading, the appropriate end date for the life of the fraud.