When the SEC announced it's financial fraud task force and a related data initiative to facilitate the identification of situations where the company “cooked the books,” many dubbed the data program “Robocop.” While the SEC may be continuing to work on that initiative, to date the agency has not created a computer to identify financial fraud. Robocop may, however, be appearing in another form. Not only does the agency use a big data approach to help sort out possible insider trading, but now it has created a program to analyze trading by investment advisers and identify wrongful conduct such as cherry picking. After test running calculations where it appeared an investment adviser had cherry picked certain option trades, a simulation designed to verify the point was run one million times. The first proceeding based the new program was filed, In the Matter of Welhouse & Associates, Inc., Adm. Proc. File No. 3-16657 (June 29, 2015).
Welhouse & Associates, Inc. is a state registered investment adviser. Respondent Mark Welhouse is its owner, principal and CCO. Custody of client accounts and assets is at a brokerage firm.
From at least February 2010, and continuing through January 2013, the firm executed trades in an S&P 500 ETF called SPY for client accounts as well as Mr. Welhouse’s personal accounts. Mr. Welhouse told the staff that he created a daily spreadsheet of the trade allocations between client accounts and his which was furnished to the brokerage firm to make the allocations. According to Mr. Welhouse, the trades were allocated on a pro rata basis before 5:00 p.m. each day.
Contrary to Mr. Welhouse’s claims, however, an analysis of all of the accounts demonstrates that in fact the allocations were not made on a pro rata basis. For trades that increased in value on the day of purchase, frequently Mr. Welhouse day-traded by selling the options on the day of purchase. A disproportionate share of the profits were then allocated to his accounts. Trades that decreased in value frequently were not sold on the day of purchase. A disproportionate share of these trades were allocated to client accounts. During this period the brokerage firm warned Mr. Welhouse several times regarding the allocations and threatened to terminate its relationship with him and his firm. Mr. Welhouse told the staff that if the allocations were disproportionate it was a mistake.
A statistical analysis of the accounts suggests that there was no mistake. During the period Mr. Welhouse allocated 496 SPY option trades to his personal accounts and 1,127 to his clients. The total cost of the trades was $7.25 million for the personal accounts and $8.46 million for the client accounts. Yet the total first-day profits for the personal accounts was $455,277 in contrast to the total first day losses for the client accounts of $427,190. Stated differently, the first day returns for the personal accounts of Mr. Welhouse was 6.28% in contrast to a -5.05% for client accounts. This compares to a return of 0.18% for all of the first day trade transactions.
The first day returns were statistically significant, according to the Order. This was verified through a simulation run, testing the possibilities. The results showed that the chance of receiving the results shown in the personal accounts was less than one in one million. The simulation was run one million times.
Finally, clients were unaware of the allocation process. Indeed, the firm’s Form ADV and related documents stated that Welhouse did not trade for its own account. The firm’s written policies and procedures stated that trades were allocated on a pro rata basis. The Order alleges violations of Exchange Act Section 10(b) and Advisers Act Sections 206(1) and 206(2). The proceeding will be set for hearing.
