Andrew Silton: Why lots of data probably won’t make you a better investor

A little over a year ago, the Supreme Court issued a unanimous decision in the case of Association for Molecular Pathology v. Myriad Genetics, ruling that isolated genes could not be patented. Prior to the Supreme Court’s decision, the breadth of genetic testing was quite limited because specific genes could be patented. For example, only the owner of the patent or a licensee on certain genes associated with breast cancer could offer the test. In the wake of the Supreme Court’s decision numerous pharmaceutical companies are now offering broad-based genetic tests that screen for mutations involving 30 or more genes at a time. What does any of this have to do with investing? Actually a great deal.

Whether it’s medicine or investing, we’re being bombarded with increasing amounts of data. At the same time that computers are sequencing the human genome, they are also devouring every possible piece of financial and market data. For patients and investors alike the great challenge is to try to figure out which pieces of data actually contain useful bits of information and which bits are noise.

While today’s new genetic tests provide a great deal more data, they are leaving us to try to figure out if a particular mutation is information requiring action or monitoring, or merely unpleasant noise. Unlike in investing, much of the noise contained in these tests will one day become useful information. I’m sure science will catch up to the power of these analyses. However, the prospects of generating more information out of financial data does not have the similarly optimistic future. Moreover, investors face an additional dilemma, which is why individual stock picking and active money management are so difficult.

Wall Street is eager to provide you with lots of data because it will encourage you to buy and sell securities. In the best of all possible worlds, at least from Wall Street’s perspective, you are required to pay for the information either through a subscription or in exchange for the fees and commissions generated by your account. Very often the information is available for free. For example, Yahoo Finance offers an almost limitless array of financial and trading statistics for anyone visiting its website. There’s only one small problem. Virtually none of the numbers, ratios, graphs, or statistics is information. It’s just a bunch of numbers or data. When you ponder a set of charts or historical data on offer from E-Trade or Fidelity, they want you to believe that you are looking at meaningful information that will help you to select a particular stock, bond or mutual fund. You probably think you’re looking at a significant pattern of historical profitability or securities prices. In reality, you are staring at statistical noise.

Predictive model needed

There are three reasons why this is so. First, most of the data that is offered is historical. Stocks, bonds and other financial assets aren’t priced based on past results. What matters is future expectations. Stock and bond prices depend on economic and industry conditions in the years ahead. Estimates of future revenues, margins and profits determine securities prices. Thus, historical data is only helpful in beginning to build a model or insight into a stock or bond. If the model relies solely on historical data, it isn’t going to work. Thus, a useful model has to have a predictive aspect, and its ability to predict has to be good enough so that there are enough winners to compensate for mistakes. Building such a model or predictive system is a daunting task.

Second, even if you have access to a predictive model or insight, whether it’s one offered by your financial adviser or built through your own ingenuity, the financial markets will have already factored your advisers or your predictions into the prices of securities before you execute your first trade. With thousands of other smart investors and traders using models and systems to make predictions, any insights offered by your or your advisers’ technique will have likely vanished before you have time to act.

Third, even if you have a model that seems to be working, it won’t be long before it stops being effective. Many of you have probably sat through a presentation in which a mutual fund manager tells you about how he has successfully purchased stocks based on a proprietary model that relies on certain financial factors, and he’s shown you return data confirming his prowess. Shortly after you make a commitment to the manager, his model stops working. It doesn’t take long for financial gurus to reverse-engineer a particular manager’s insight and render it ineffective.

Financial markets learn very quickly and adapt to anyone who tries to apply the same financial insight repeatedly. It reminds me of what happens when a particular antibiotic is overprescribed. The bacteria, much like the financial markets, mutate, and the antibiotic becomes useless.

Data plentiful, info scarce

Perhaps, the scarcity of information in a sea of data explains why some investors are tempted to engage in insider trading, and why high-speed trading is so attractive. In both cases – one illegal and the other legal – the investor has an edge that the market can’t easily eliminate because the information can’t become mere noise. Insider trading relies on information that is not yet known to the market. Thus, the price of a particular security cannot reflect either the good or bad news about a pending earnings report, product introduction or merger. Armed with inside information, a trader can confidently buy or short a stock, knowing that the information will eventually hit the market and vindicate his position. Since the legal means of extracting meaningful market information are so rare and fleeting, there’s a huge temptation to look for and act on information that’s not yet publicly available. I’m quite certain that future generations of prosecutors will indict future generations of money managers seeking the inside trader’s edge.

High-speed trading is alluring because it doesn’t rely on fundamental information, or the ability of the markets to factor in that information. Rather, it depends on having enough speed to jump ahead of the trades of slow pokes like you, me and most institutional investors. Even if our trades are executed in less than a second, that’s lethargic compared to high-speed trading. Using ultra-fast computer and fiber-optic cables, the high-speed trader can “see” and act on supply or demand, building in a particular stock before anyone else. Imagine someone being so fast that they’re always able to cut into the front of the line for a popular movie without anyone noticing; that’s high-speed trading. Setting up a high-speed trading network involves huge costs. Thus, the ability for someone else to build an even faster system and eliminate the high-speed trader’s edge is limited.

Since most of you aren’t inclined to break the law and don’t yet have the money to retire, let alone build a high-speed trading platform, you’re left with a choice. You can spend your evenings sifting through the mountains of data looking for an investment edge or hire a manager to undertake this difficult cause. However, almost every other element of your life from health care to education seems to be built on growing mountains of data. In investing, there is a way to cut down on the noise, generate reasonable returns, and concentrate on the more meaningful elements of life: index.

Andrew Silton’s Meditations on Money columns can be found twice a month in The N&O’s Work&Money section. He is a retired money manager living in Chapel Hill. He was CIO for the North Carolina Retirement System from 2002-2005. He writes the blog