Sorry we’re late this week! We’re in Las Vegas and the Market Structure Map wanted to stay in Vegas, apparently.
Program trading jumped about 10% in the first few trading days of April, compared to March. We define program trades as mathematical execution in more than one stock at a time that follows a market trend. We believe the increase in part reflects improved institutional and retail commitments to various equity index vehicles. But it also shows how “market neutral” trading strategies that continuously buy and sell a wide variety of securities to achieve small profits when netted out have become mainstream.
It shows up in interesting ways. For instance, a client of ours in the telecommunications sector with market capitalization of about $500 million is in nearly the same number of ETFs – 71 of them – as a network storage technology company with market cap well over $30 billion. It illustrates that small, rapid trades in many securities is a successful and profitable way for market participants to generate returns. Algorithms might incorporate a basket of ETFs, a set of interest-rate futures, some bonds, and a whole index of stocks and juggle them throughout a day. If the trading desk is a “liquidity provider,” it probably pays little for the trades and may in fact turn a profit on the trades themselves, regardless of securities gains. At the end of a trading day, with puts and takes, the algorithm might return a few basis points. Done over and over, it’s a low-risk way to make very good money.
From trading firms like Hudson River, to the prime brokerage desk at RBC Capital, this is the way things are done today. It’s a serious challenge for conventional surveillance to tackle because what shows up in settlement data isn’t what’s occurring in the market and setting price. So you may be telling your management company that your stock is up on buying at Southeastern Management when in fact it’s proprietary trading in beta portfolios.
The key is to know the difference. Speaking of which, we promised regular real-world examples from market-structure analytics. Here’s this week’s, on setting expectations for price performance ahead of earnings. A client of ours providing green technology solutions, a small firm with market cap just over $100 million, was set to report in early March. We could see that what we call “rational price,” or the price where active money steps into the algorithmic rhythm of the market, was substantially lower than where the stock was trading. We told them where their price would go if they met, beat or missed. They missed and the price dropped and stayed at the price we’d indicated, to the very penny. Until the day the first bit of value money returned (which we also could see). We’re not always that precise, but we can almost always see within a percentage point or two, because we’re tracking behavior, not trend lines.
The beauty here is that no matter how much algorithmic rhythm your stock may have as it moves through 100 different ETFs, the smallest footprint of real buying is still the greatest price denominator. IR matters. But unless you know what’s behind your price and volume, it’s hard to prove.
