August 31, 2010

Missing the Mark in Algorithmic Trading

Do you think your stock trades well?

While you ponder, a confession: We’re guilty of a bait and switch. If I’d written “implementation shortfall,” which is what I mean, rather than “missing the mark” above, which is what I said, I might be responsible for a chain-reaction narcoleptic catastrophe, people randomly falling asleep mid-word and banging heads on laptops, iPads, desks, afternoon pub beverages.

Implementation Shortfall” is some software engineer’s term for not meeting your target in a trading scheme. By analogy, say you want to sell your house for $500,000 and you get $480,000. That’s implementation shortfall. In trading, it’s the difference between what the client wants and what you deliver. Say you’re tasked with selling 200,000 shares within a price range of $0.25, and with set trading costs. The degree to which you don’t meet that target is implementation shortfall.

In the IR chair, you don’t need to know all that stuff. But you should understand how implementation shortfall can impact how your stock trades. Maybe you can’t target the institutions you thought you could because at the trading desk, the implementation shortfall risk keeps trades from occurring. Plus, it’ll show up in your stock price. I’ve got a story to tell to illustrate it.

But before that, a word on trading since options expirations 8/18-20. The Dow Jones Industrial Average is off 440 points since. Data for clients show the chief cause not to be a rout from equities by informed money, but a shifting of the balance of resources. An institution at expirations decides to swap $500 million of tech stocks for a basket of bonds and Treasuries. The transaction does not hit the open market, but because the investor had been trading around his position before expirations and now has ceased, the effect is a gradual decline in the whole sector, because each algorithmic trade causes a ripple into the behaviors of other algorithms, many of which are there to match up trades now occurring at a lower level. The sector slides 10%.

It relates to algorithmic (mathematically managed) execution. If the placing of trades is less than random, machines identify an absence of randomness and game the order flow. What we have at the moment instead is a widespread effort to randomize so others cannot front-run orders and cause implementation shortfall. The result is chaos, combined with small shifts that produce big changes in price.

In markets where investment behavior, or rational thought, is principal price setter, value money would be at work, seeking value. We don’t have that. Instead, money seeks alpha or reacts to change. That’s why you see randomness everyday in trading.

Which brings us to our story. A particular basic materials small cap stock showed miniscule nudges in algorithmic trading on but two of ten program desks. Yet this smallest lack of randomness between overall buying and overall selling caused slippage of nearly 20% in the price. No value money tried to buy shares until the whole implementation shortfall had run its course by August 26. Perhaps because it too had been randomized. Unless you see these features in your own trading, you can’t truly say if your stock trades well or not (it varies, by the way, which is why IR must learn to be tactically nimble).

One answer to this problem is for IR professionals to color a broader palette of institutional relationships, not just buy-and-hold sorts, who have largely gone random. I was speaking with Jim MacGregor at IR/PR consultancy MacGregor Abernathy, and I agree with Jim that market structure almost demands that IROs court a coterie of aggressive hedge funds.

In a sense, hedge funds are the last institutional investors standing whose actions haven’t been randomized and ionized into a form of managed, program-driven behavior incapable of responding to value opportunities. I’m talking of course about active hedge fund portfolios, and not, say, DE Shaw’s Oculus portfolio. DE Shaw runs active money too. Their ability to act can alter the randomness in your market structure with one small hedgy step.

And now if you’ll excuse me, my schedule is experiencing an implementation shortfall.

Share this article:

More posts

dreamstime m 206876446
February 14, 2024

I’ve got my Valentine, for which I’m grateful every day.  Whether the market finds love after yesterday’s blood remains to be seen. Back when the...

dreamstime m 212403964
February 7, 2024

On Nov 15, 2021, NVDA closed at $345.30 on a hundred million shares of volume. Without context, that information is interesting but unhelpful.  I’d note...

dreamstime l 24803177
January 31, 2024

Consumers are confident. I’m not sure they have all the data. It’s a lesson for public companies. In case you missed it, The Conference Board’s...

dreamstime l 56087804
January 24, 2024

If I said the name “Sherlock Holmes” to you, what’s your snap response? Probably, “Elementary, my dear Watson.” I have long favored a line by...

dreamstime l 50373701
January 17, 2024

What do bitcoin Exchange Traded Funds mean for public companies?   More competition for those dollars you chase with your story. A number of you...

dreamstime m 4397737
January 10, 2024

Total nonfarm payroll employment increased by 216,000 in December. So said the headline from the Bureau of Labor Statistics in its monthly report on jobs...