Last week, markets were abuzz over the zero that lifted LYFT.
For those who were vacationing or living hermetically and missed it, Lyft reported financial results and inadvertently added a zero to adjusted EBITDA margin-expansion, giving machines the notion profit margins improved 5% instead of 0.5%.
It was a great Valentine’s Day for LYFT investors as the stock surged 35%.
Illustration 20057394 © Benchart | Dreamstime.com
LYFT issued a correction immediately. Read it here. Yet the stock not only held those gains but added to them the next day, rising another 16%, for a two-day gain of 57% on an errant and retracted zero.
Maybe results were just good, Tim?
Maybe. But Active money didn’t boost shares 57%. On Feb 13, Active Investment was 12% of LYFT trading volume, and by Feb 15, it was 10%, down 19%.
But quantitative volume and trades tied to derivatives like call options rose a combined 150%. There was an error. The stock shot up on quant flows. The error was changed. The stock shot up further on quant flows.
To explain, “quants and derivatives” reflect trend-following investment tactics where algorithms buy both underlying stocks and the associated call or put options to magnify short-term moves. Options can boost returns by factors, not percentage points.
What should we learn from LYFT’s gain?
For one, fixing the error was immaterial. That’s a deduction derived from how the price behaved and what the data showed.
And we can infer that prices themselves are more powerful than the data that created the prices. How so? The prices remained after the data error was fixed, and higher prices were added atop those prices.
Machines don’t know to back out incorrect prices. In fact, machines don’t differentiate between “correct” and “incorrect.”
Imagine the moral hazard here. Why not…add a zero someplace? And then reissue a corrected release? I hope not. But consider human nature.
Which manifestly leads to the Big Conclusion: Machines react to data. Not its rectitude.
TOST was up 17% on earnings. So far as I know, there were no rogue zeros. Driving price up was, you guessed it, a 62% jump in the same stuff lifting LYFT: Combined quant volume and derivatives bets.
EPAM was up 12% the two days after earnings on a 45% jump in quants, derivatives, as a proportion of daily trading volume versus the day before earnings.
ROKU was down 24% on 34% higher combined quant and derivatives trades as a share of trading volume, day-over-day. ROKU has a very short release. But ROKU was 67% short – over two-thirds of its volume was borrowed, created, artificial – before earnings. Quants were loaded up on short shares and put options before the release hit.
Opendoor was down 10% on a 101% explosion in quant and derivative volumes.
MSTR was up 55% by Feb 15 after earnings Feb 6, on a 25% sustained increase in buying of stock and derivatives by quants, while volume from stock-pickers declined 1%.
I could go on and on.
Almost never do stock-pickers juice ups or downs on results. One exception, The Trade Desk (TTD) was up 17% with earnings on a 47% increase in Active Investment. But quants and derivatives were up 52%, still more than stock pickers.
Machines don’t react merely to errors. Machines react to data. And to prices. The data in your press release is ammunition for trading machines, errors or no. Machines are machine-reading your release, public companies, and they’re trying to foster pricing inefficiencies so they can profit on volatility.
For nearly 20 years, the SEC has approved every form of derivative, from 0DTE options to leveraged long/short ETFs. And Active managers have lost trillions of dollars to Passives. Passive investment is now 70% of equity assets at the seven largest US money managers.
Public companies, we’re our own worst enemy. Big Passive money wants a stable, low-volatility product that will predictably track a benchmark. And every quarter, public companies stuff earnings releases full of data points, like clips of machine-gun ammunition, that rips and tears through the stability Passives seek.
And this is how 72% of the S&P 500 underperforms its own benchmark.
Passives don’t have to own you. At least not often. Rules give Passives roughly 70 days after quarter-end to true up tracking. Most Passives use a statistical sample of an index or benchmark as “the basket.” That group will be large, stable, predictable.
To remain in it, you need to be large, stable, predictable. You need to be the herd. Not an outlier. Errant zeros aren’t volatility requisites. Any data point will do.
So make them few, public companies. Change your practices. Gone are the days of outliers, the lone elk breaking from the herd. Be the herd. Neutralize your earnings release. Disarm the machine gun.
Because to trading machines, you’re ones and zeros.
