Saturday, 30 May 2015
Last updated 11 hours ago
Nov 30 2009 | 9:10am ET
By Jeffrey Rathgeber -- In Part I, we discussed how to structure your methods to get the permissions and access you need to screen for hedge fund insider trading.
In Part 2, we go over how to do some basic screening for insider trading as part of hedge fund due diligence.
As always, the trick is to not over-complicate things. By constructing screens, you're putting yourself in a position to gain cold comfort that large-scale and repetitive insider trading is not the manager's core strategy. There is no foolproof method to detect insider trading, but you can pick-off the low-hanging fruit in a short amount of time if you arrive prepared. Below are three of our favorite screens.
First screen: The Lottery Ticket. We're talking about large single-day/single ticker gains in a sector clearly outside the manager's specialty. Most managers rose up in the business by being very good at trading one or two sectors. There are very few generalists in the business. And even in multi-strategy shops, each portfolio manager usually has a specialty. The P&L records always have a way of separating the performance of these PMs. Armed only with intermediate Excel skills (i.e. - pivot tables), a due diligence pro can quickly isolate the energy PM who made a one-shot killing in Sony or Pepsi. Again, any one stand-alone instance like this is proof of nothing, what you're hunting for is patterns. When the energy guy's five best trades from that period are in retail, technology, healthcare, healthcare, and healthcare...caveat emptor.
Second screen: Halted Stocks. This is about large single-day/single-ticker gains in halted stocks. Bloomberg access is required for this one, but all fund administrators have Bloomberg and you're sitting in the fund administrator's offices when conducting this part of diligence, so problem solved. Have the fund administrator download all trading activity for the period under examination. Use that basic Bloomberg function all administrators use to download closing prices directly into a spreadsheet right next to a column of tickers. Except, instead of pulling the closing price from Bloomberg, use the heading "HALT" which will return a simple Y or N depending on whether the stock has been halted. Isolate the Ys and if it's there, it will be painfully obvious. You'll see time and price of the trade (normally an enormous short position) put on minutes before the Halt time. Case closed.
Third screen: Magic Crystal Ball. So what if a PM makes a killing inside the sector he normally trades and nothing was halted in the process? Again, there's either a pattern or there isn't. If the top five single-day/single-ticker gains were positions that were initially put on (or dramatically increased in size) shortly before news came out that either spiked or caved the price, then you can either believe that the manager’s crystal ball is actually that good, or you can diagnose that a pattern exists that is clearly not repeatable which plays a significant role in their returns. We're being kind by avoiding words like illegal, immoral and unethical, but I think illegal should be enough.
The mechanics of filtering massive amounts of data honestly requires little more than mid-level spreadsheet skills. In this age of massive frauds and colossal failures of honest disclosure by blue-chip managers, nobody is above these questions and methods, and no fiduciary can allow the money of others to sit with managers who say they are.
Jeffrey Rathgeber is a founding partner of Pelorus Advisors, a leading authority on Capital Risk Management. Rathgeber is an experienced hedge fund CFO and runs the firm's Hedge Fund Investor Assurance practice group. Pelorus supports leading investors in the alternatives space, including pensions, family offices, endowments, and fund of funds.
May 27 2015 | 2:15pm ET
Support Hedge Funds Care, also known as Help For Children (HFC), by participating in this year's raffle. All proceeds go to support HFC's mission of preventing and treating child abuse. Read more…