FactSet says quarterly earnings are up 23% from a year ago. Why have stocks declined?
There’s an inclination to grasp at fundamental explanations. Yet stock pickers generally don’t reactively sell because most times they must be fully invested (meaning to sell, they must buy).
Blackrock, Vanguard and State Street claim for Exchange-Traded Funds tracking the S&P 500 or Russell 1000 that turnover is 3-5%. (Editorial note: Those figures exclude creations and redemptions of ETF shares totaling trillions annually – a story we’ve told exclusively in the Market Structure Map.)
If investors are not responsible, who or what is? Machines. By market rule all trades wanting to set the best bid to buy or offer to sell are automated – running on an algorithm. Why? Because the best price can be anyplace at anytime in the market system, and trades must move fluidly to it.
Thus, machines have become hugely influential in determining how prices are calculated. An amalgam of broker algorithms, smart routers and exchange order types are continually calculating the probability of higher or lower prices and completing a trade.
By our measures, back on Apr 19 the probability of calculating higher prices dropped. Why? Perhaps risk calculations for asset managers ordered rotation from overweighted equities or a need to slough off capital gains from ETFs (stuff mathematical models routinely do).
We have a mathematical representation for it: The market was Overbought. It doesn’t mean people are overpaying for fundamentals. It says machines will lack data to arrive at higher prices. What follows this condition is nearly always a flat or lower market.
We know then that math arising from market rules is more powerful than a 23% increase in earnings. That should disturb stock pickers and public companies. If the market is The Matrix (if you’re younger than the movie, watch it to understand the reference), what are we all doing straining so hard to be outliers?
And why do machines possess the capacity to trump value-creation?
Good question.
By the way, the math is now changing. It’s resolving toward a mean. We measure these price-setting propensities with a 10-point scale, the ModernIR Behavioral Index. Most of the time the stock market trades between 4.0 and 6.0, mean-reverting to 5.0 or thereabouts.
It returns to the middle because rules propel it there. Stocks must trade between the best bid or offer. What lies there? The average price. What do indexes and ETFs hew to? Averages. We’ve explained this before.
When the market slops beyond 6.0, a mean-reversion is coming. When it drops below 4.0, it signals upward mean-reversion. The market has descended from about 6.5 a week ago to 5.2 yesterday. The market will soon level off or rise as it did microcosmically yesterday, a day of extremes that ended back near midway (but it’s not down to 4.0, notice).
If math is a more reliable indicator of the future than earnings, why is everybody fixated on earnings versus expectations? What if that model is obsolete? And is that a bad thing?
I don’t think so. The earnings-versus-expectations convention promotes arbitrage. Shouldn’t capital-formation power the market?
