In late December, Jeffry Picower announced that his Picower Foundation was closing because it lost a billion dollars in the Madoff scandal. In May, Irving Picard, the Madoff bankruptcy trustee, sued Picower for $5.1 billion, claiming Picower knew transactions in the BMIS accounts under his control were fraudulent.

A review of the foundation's 990s from 2001 through 2007 show unusual transactions that Picower would have known were bogus. The question is whether the foundation's independent trustees should have realized something was amiss.

One of the trustees, William D. Zabel, is Picower's long time lawyer and spokesperson. Zabel is a partner in the firm of Schulte Roth & Zabel in NY. 

Another trustee, Gerald C. McNamara, I believe, is a managing director at Goldman Sachs.

The other two trustees are Norman B. Leventhal, a Boston real estate developer and Martin Post, a cardiologist.

The Picower Foundation did not pay an outside accounting firm to prepare the 990s. In 2007, accounting fees were $1,800 so it is unlikely the foundation's books were audited. In 2005 and 2006, the fees were only $1,250. The fees were paid to a small firm in New Jersey.

In 2002, 2003 and 2004, the foundation paid accounting fees of $14,000, $21,500 and $15,500, respectively, to KPMG, a "big four" firm.   

The foundation's stock portfolio appreciated remarkably over the years. For example, the foundation purchased four stocks on 7/14/05 for $141 million which resulted in a net unrealized gain of $62 million within six months.

There were also discrepancies in the portfolio from year to year. In 2005, for example, the foundation held 174,000 shares of  Research In Motion, purchased on 5/3/04 for $44 per share for a total of  $7.7 million. In 2006, the foundation claimed it held 384,000 shares purchased on 5/3/04 for $62.36 per share for a total of $24 million.   

Here's a partial snapshot of the foundation's stock portfolio at 12/31/07 included in the 2007 990:

 
The foundation had $641 million in stocks at fair market value and $312 million in treasury bills at 12/31/07.

Bernie's split-strike conversion strategy obviously was not being employed since he always converted to treasury bills at year end and he generally did not hold stocks longer than a few months.

In 2007, the foundation sold $53 million of stocks for a net gain of $31 million. 

The 2006 portfolio included in the 2006 990:

The foundation had $432 million in stocks at fair market value and $174 million in treasury bills at 12/31/06.

In 2006, the foundation sold $133 million of stocks for a net gain of $109 million.

Compare the data for the Research in Motion shares in 2006 to 2005. As noted above, the number of shares purchased on 5/3/04 changed as did the share price.

In 2005, the foundation held no shares of Oracle in 2005 but, in 2006, it reported holding Oracles shares purchased for $8 million on 12/3/01.  

The 2005 portfolio included in the 2005 990:

The foundation had $418 million in stocks at fair market value and $175 million in treasury stocks.

On 7/14/05, the foundation purchased stock in Amazon, Apple Computer, Burlington Reources and Google for $141 million. By 12/31/06, the value of the four stocks had increased to $203 million. These were the only stocks purchased in 2005.

The foundation did not report any sales of stocks or other assets in  2005, 2004 and 2003.

In 2005, Eagle Materials shares were listed as having been purchased in 2001, however there is no record of the stock in the 2001, 2002 and 2003 990s.

What Bernie's strategy was supposed to be is not evident. Stocks were purchased and sold sporadically and sometimes held for years unlike the other Madoff accounts. 

In December, Zabel made a point of telling the New York Times that the investments were virtually entirely handled by Bernie. We now know that wasn't true. But did the trustees really believe Jeffry Picower, a sophisticated investor, just handed Bernie the reins?

Did anyone wonder how Bernie was so successful at picking stocks? Did McNamara, a Goldman Sachs director, ever question the results?