How To Invest

Want to know more?

Find out more arrow-white

Island Investing

Riffs, rants, and the upside of investing from way off Wall Street


How Should Long-Term Investors Think About High Frequency Trading?

From my Q4 letter to Tarpon Folio investors in January:

“Never Mind BitCoin: Stock Market Already Chock Full of Shady Algorithms”

Daily trading in the U.S. stock market has become dominated by computers. Some estimates put algorithmic or “algo” trading at close to 80% of daily volume. Assuming that’s accurate, then the best indicator of the algos’ impact on the market every day is the S&P 500 Index itself. In other words, the “market” has become the algos, and vice versa. So it should follow that in order to beat the index, you’ve got to beat the algos.

Now, if I asked you to describe the best way to compete against a thousand racks of the highest-powered servers in the world, running code written by the kind of folks who thought the large Hadron collider wasn’t a challenging enough project…well, let’s be honest, not many of you would say:

“Yeah, find me someone on a remote island chain. Someplace with lots of rum and rusty fish hooks lyin’ around. Maybe some drag queens nearby. I want this guy to invest my money using a process that started in the 1930’s. He’s gotta spend most days alone, reading stuff, from inside someplace dank. You know, like an old explosives garage. But throw a dog in there with him so he doesn’t go all Walter White.”

Here’s the thing, though: we’ve beaten the machines over the past five years. Not even close, actually. So I ask you – which way to invest is insane, really?

If you’re competing with the algos, you need a process which gives you the best odds of coming out on top over time – because the shorter your time frame, the less likely you are to come out ahead. Besides, focusing too much on the daily noise means you risk short-circuiting the hard drive in your own head.

Because these algos can process and trade on new financial data within seconds of it being published, you should also realize the futility of attempting to beat the market by basing your investment decisions solely on things that are easily quantifiable by a computer. To win, you need to pick your spots, focus your efforts, go where the algos cannot or will not go and/or have as core to your investment thesis something that cannot be easily captured in a programmer’s code.

So for us, I think the main thing to keep in mind when it comes to 2014 is this:

Myopic machines do not care about the absolutes of over- or under-valuation when it comes to a company’s shares…only ‘better’ or ‘worse’ than prior expectations.

Most algos focus on changes to a company’s expectations relative to just thirty seconds ago. And many high frequency trading algos appear to exist solely to flood the exchanges with fake orders in an attempt to scalp shares to the next investors who show up – as if they were desperate tourists outside Madison Square Garden.

To be blunt, I do not care for high frequency trading. Legalities aside, it’s effectively a $10B-to-$20B-a-year tax being paid by unsuspecting investors – and the megaminds who collect don’t even bother to come pick up the garbage on Tuesdays. That said, you’re not paying me for expertise in municipal waste management.

The reality is that when it comes to the long-term performance of Tarpon, these algos aren’t necessarily a bad thing. In fact, they can be very good.

More specifically, if a stock’s price is driven disproportionately further below its true intrinsic value due to short-sighted algorithms (or the emotional investors they fleece), then as long as we can stomach the volatility, we should eventually realize above average returns.

That does come at a price, though – a temporary emotional one, really, which I think means it’s all the more important that our investing philosophy be grounded in reason. We will “be early,” or buy stocks as they are still going down. Some of the companies we own may induce nausea at first glance. Others may be small – too small for algos to trade in. All of those are just part of the deal in Tarpon.

The presence of a “catalyst” – a positive future event that a programmer can’t confidently model – has also become that much more important to the case for investing in a given company of late.

So let’s set aside prognostications about the Fed, interest rates and inflation. When it comes to Tarpon specifically, my crystal ball for 2014 looks a lot like it did at the beginning of each of the last six years. It’s a bit fuzzy, but I foresee lots of time in some kind of bunker. In there with me is either a very large ferret or my black lab, not clear yet. Either way, I am hunkered down and attempting to correctly anticipate significant positive revisions to the long-term expectations for the best, most undervalued companies I can find…so we can profit accordingly over the coming years.

To receive my letters, please sign up here.