A challenge to the Commissioner's use of the net worth method of determining income is presented by this petition seeking review of a decision of the Tax Court of the United States.

Petitioner George Schwarzkopf, an oculist and eye surgeon, was a resident of Atlantic City, New Jersey, during 1947 through 1951, the taxable years under review. His residence contained not only his living quarters but also his office, operating room, and hospital rooms. Taxpayer kept his own receipts book from which he reported his yearly income. He admitted that many of the sums entered on his receipts book represented not gross but net fees, after deducting certain hospitalization expenses. Taxpayer did not show any of the receipts from these expenses to the investigating agents, and most of the receipts from earlier years were admittedly destroyed.

Finding a substantial discrepancy between bank deposits and reported income, the Commissioner reconstructed the income of Schwarzkopf by resort to the net worth method, with the following comparative results:

                                   Unreported
                  Net income as  taxable income
      Net income  determined by  determined by
Year   reported    respondent      respondent
1946  $14,688.25   $ 49,140.24    $ 34,451.99
1947   14,244.31     31,908.91      17,664.60
1948   14,200.00     50,744.70      37,544.30
1949   16,040.59     34,376.53      19,335.94
1950   17,000.79     48,125.46      31,124.67
1951   21,721.06    175,466.96     153,745.90

The Commissioner then determined the taxpayer's deficiencies for the respective years and additions to tax under sections 293(b)1 and 294(d)(2)2 of the Internal Revenue Code of 1939 as follows:

                        Additions to tax
Year  Deficiency   § 293 (b)   § 294 (d) (2)
1946  $ 20,027.69  $10,208.94       $1,322.46
1947     9,356.04    4,678.02          545.66
1948    19,942.57    9,971.28        1,157.59
1949     9,301.21    4,650.60          586.66
1950    17,020.85    8,510.42          984.73
1951   125,811.42   62,905.71        7,710.98

Most of the amounts utilized in the Commissioner's net worth analysis were stipulated. Only undeposited cash on hand was disputed, and the Commissioner determined that there was none as of January 1, 1946 and $32,000 undeposited cash as of December 31, 1951; these amounts were prorated over the years in question. The Tax Court, however, found that taxpayer had $101,000 in undeposited cash as of January 1, 1946, and the following amounts thereafter:

December 31, 1946  $62,000
December 31, 1947   52,000
December 31, 1948   32,000
December 31, 1949   32,000
December 31, 1950   32,000
December 31, 1951   32,000

The Tax Court determination, of course, completely eliminated the 1946 deficiency and made certain changes in the deficiencies of the other years involved. Finally, the Tax Court found that petitioner's income tax returns for 1947 through 1951 were false and fraudulent and were made with intent to evade tax, and that part of the deficiency for each of the years 1947 through 1951 was due to fraud with intent to evade tax.

The salient thrust of petitioner's principal argument is that the Commissioner's use of the net worth method of income reconstruction was unwarranted as his books of account were correct and adequate.

Holland v. United States, 1954, 348 U.S. 121, 75 S. Ct. 127, 99 L. Ed. 150, has become the fons et origo of the law concerning the use of the net worth method of income reconstruction. In giving its approval to the net worth method, the Supreme Court said, 348 U.S. at page 132, 75 S.Ct. at page 134:

'* * * To protect the revenue from those who do not 'render true accounts,' the Government must be free to use all legal evidence available to it in determining whether the story told by the taxpayer's books accurately reflects his financial history.'

This quoted portion of the Holland case is recognized by petitioner as sanctioning the use of the net worth method to test the accuracy and completeness of the books of account. Thus, the net worth method serves two purposes: first, it may be used to test the correctness of the books; secondly, it is cogent evidence of the amount of income which went unreported. The fact that the books on their face appear to be adequate does not preclude the use of the net worth method. Holland v. United States, supra, 348 U.S. at pages 131-132, 75 S.Ct. at page 133. In any event, the books involved here contained items of net income with hospital expenses already deducted. The disposal by the taxpayer of bills evidencing these expenses made the computation of their amount impossible, and thus left vague and unreported some unknown amount of income. The taxpayer's practice of cashing checks representing his patients' fees and receiving the money in large denominations rather than depositing the checks themselves made it impossible to test the accuracy of the books from that source. If taxpayer's contention is correct, everyone could keep a set of apparently accurate books, carefully destroy other evidences of the source and amount of income, and defend by an alien rule that the net worth method may not be used in those circumstances-- and thus the government could be defrauded with impunity. However, it is when other methods of disclosing income fail that the net worth computation becomes especially important in the collection of revenue.

A favorite defense in net worth cases is that taxpayer had a large amount of undeposited cash on hand at the beginning of the investigation period. This may explain the disproportionate increase in net worth over the increase of taxable income. Petitioner here likewise contends that he had on hand a sizeable amount of undeposited cash at the beginning of the investigation period. Evidence as to the amount and source of this cash is questionable at best. Petitioner contends that the Tax Court committed clear error in refusing to find that a certain large sum came to him by way of an inheritance from an uncle who died in a land now occupied by Russia. Documentary proof of death and of the inheritance were not available because the records there were destroyed, and the only evidence of the inheritance were self-serving notations on the taxpayer's passports of the amounts he allegedly brought back from Europe. The fact that the Commissioner did not directly contradict this evidence would not compel the Tax Court to accept it in the circumstances of this case. The Tax Court, evaluating witness credibility, simply did not believe the story.3 We think it was eminently proper to use the net worth method here.

The burden of proving fraud is, of course, upon respondent. Section 1112 of the Internal Revenue Code of 1939, 26 U.S.C., 1952 ed., § 1112. Petitioner contends that respondent failed to sustain his burden. This point shall not long detain us. Even if this case were devoid of the usual indicia of fraud, the consistent failure to report substantial amounts of income over a number of years, stating alone, is effective evidence of fraudulent intent. See Holland v. United States, supra, 348 U.S. at page 139, 75 S.Ct. at page 137; Lipsitz v. Commissioner, 4 Cir., 1955, 220 F.2d 871. Over a period of five years, petitioner failed to report over $200,000 of taxable income.

Finally, petitioner argues that he is not liable for penalties for underestimation of tax based upon Section 294(d)(2) of the 1939 Code for the years 1948, 1949, and 1951. We agree with petitioner's contention. The section in issue provides an exception from its penalty provisions for those years in which there has been a timely payment of estimated tax in an amount at least as great as it was 'on the basis of the facts shown on his return for the preceding taxable year.' Petitioner's estimated taxes for 1948, 1949, and 1951 were $4,000, $3,600, and $4,000 respectively. Each amount was greater than the amount of tax shown on taxpayer's returns for the preceding taxable years. If, as the Commissioner contends, Congress meant to compare the estimated tax with the tax based upon the reconstructed income, it should have so provided. The penalties under Section 294(d)(2) should not have been assessed for 1948, 1949, and 1951.

The cause will be remanded to the Tax Court for an adjustment of the penalties under Section 294(d)(2) and, as adjusted in accordance with this opinion, the decision will be affirmed.