Tagged: AI

Panakes

Posted on June 21, 2023

The Latin word Panacea comes from the Greek “Panakes,” a combination of two words meaning “all healing.”

I love the Greeks.

Humans love panaceas.

Photo 23332434 | Panacea © Feverpitched | Dreamstime.com

Like central banks. Fat pills.  Digital assets like bitcoin. The metaverse. Dutch tulips. Vaccines. A silver bullet. Artificial intelligence.

We humans latch onto things in hopes they’ll solve all our problems. 

There’s nothing wrong with hope.  It’s quintessentially human.  But hope isn’t a strategy and sometimes it obscures reality.

A WSJ article yesterday (subscription required) tries to explain the rally in Tesla’s stock under a title that, dissonantly, says it doesn’t make sense. It’s because Tesla is an AI play. It’s charging stations. It’s robotaxis. Musk appointed a Twitter CEO so he can focus.

And so on.

Not a single sentence reflects an understanding of the ecosystem for TSLA shares.

This same condition often manifests amid the human propensity to embrace panaceas.

Underneath Tesla’s lofty valuation are rational things, yes.  And it’s a big business with 100,000 employees.  But the ecosystem that sets TSLA’s price treats equities like products, not nuanced stories.

TSLA is above all else a Large Cap Growth stock. Of the 20 largest ETF positions – TSLA is in 325 – eighteen are large cap growth or total market. The other two are Consumer Discretionary.

So, what’s behind TSLA’s valuation is, simply, money wanting a large cap growth Consumer Discretionary product. 

TSLA and NVDA, just the two, are 10% of total market dollar-volume. Add the other five of the so-called Magnificent Seven (we have no acronym now) to them, and they’re 30% of the S&P 500.

And 98% of Large Cap Growth stock-pickers don’t beat the benchmark. Do you see the problem?  The WSJ article is built on a series of fundamental factors. And the author says they don’t explain the rise.

Correct.  What explains the rise is the ecosystem. Not a panacea like AI.  Big money needs big stocks.  There aren’t very many. So the money is concentrated ever more into the few.

NVDA now trades over $40 billion of stock daily.  The entire stock market trades about $600 billion daily.

It’s a Demand/Supply question. An ecosystem question.  That will become an ecosystem problem at some point.

Nearly all the $45 trillion market capitalization in US stocks is in 800 stocks, meaning some 3,000 small caps and micro caps are left out, adrift.

Shouldn’t companies rightly and fairly know the odds before they IPO?

My profession, investor relations, is also hoping for an AI panacea. There’s a post in my profession’s web forum where ideas are genially and generously shared about wouldn’t it be wonderful if AI powered a perfect dissection of investor data so we could precisely target investors right for us?

A panacea.  Finally, a way to address the troubling absence of investor flows.

This post came from a small-cap company, one of the three thousand adrift on an endless sea surrounding the seven islands that drive its currents.

The trouble is there’s not enough money actively targeting stocks to move the shareholder-value needle.  The 10% of volume that stock-pickers drive is overrun, overpowered, by the machines – that are already using generative, pre-trained transformation to mash up statistics and execute trades with algorithms.

No amount of careful targeting can surmount the obstacles presented by the ecosystem: The money – over 40% of all institutional dollars now – is Passive Large Cap Blend.

So what is the panacea?

I’d argue that panaceas are the problem. Panaceas breed groupthink. Which promotes concentration. Which creates bubbles. Which destroy confidence. Wash, rinse, repeat. 

Everything cannot be healthy.  Just a fact.

What’s the message of hope? For investor relations, it’s shifting strategic focus from promoting the story to feeding your characteristics.  What’s your strategy to become a Passive Large Cap Blend stock?  You have only two choices: Deals, catalysts.

So understand what kind of product your shares are for Passive money.

And streamline what you do. Calls, sellside conferences, investor outreach.  If you want to know more, we can explain. Actions should produce outcomes or be altered.

Investors, this is not Peter Lynch’s market. You can’t uncover undiscovered stories. There were 8,000 public companies then, and Intel and Microsoft could start small and grow big.

Today, there are 3,500 in the national market system, and 2,500 of those are 5% of market cap and just 1% of those will achieve escape velocity. Not even the craziest gambler bets on 1% odds.

Machines already know all knowables. You can only buy and sell the same stocks everybody else owns – at different times.  Try Market Structure EDGE. There is enormous alpha in Demand/Supply imbalances. You just have to know how to see it.

AI is no panacea for stocks, investing, investor-relations. The ecosystem determines outcomes.

To change that, we’ll have to shake up the ecosystem. That’s the next big challenge for capital-markets professionals. And that’s real rather than artificial intelligence.

Ford Pinto

Posted on April 26, 2023

Remember the Ford Pinto? 

(EDITORIAL NOTE: also remember to join tomorrow’s weekly live Demo and discussion for Market Structure EDGE, if spots are still available).

Photo 18334174 / Microsoft © Jakub Jirsak | Dreamstime.com

If you were born in the 1980s or later, you probably don’t know the Pinto.  But thanks to internet search engines, you can look it up.  There’s a scene from the cheesy Val Kilmer comedy Top Secret! where a German military vehicle taps a Pinto bumper and blows up.

In a sense, FRC (the bank, First Republic Corp) tapped the bumper of the stock market and it rang like crystal and then detonated.  I’m joking, mostly.

I’ll come to that.

A bit more history:  In 1998, I was working for a startup (where I rose from sales guy to President) and we shifted from typing stuff into a new outfit called Ask Jeeves to search for answers and started using a whizbang thing called “Google.” 

We said amongst ourselves, “This could catch on.”  

But did we really think in 2023 that Bill Gates’s outfit, which was a big deal in 1998, would be the one to ford Google’s moat and threaten the fortress?

We looked at MSFT and GOOG market structure Monday on Benzinga Premarket Prep (I do it most Mondays, 830a ET) and I noted the superiority of MSFT’s Suppyly/Demand balance. Steady – solid, we could call it – Demand, 30-day downtrend in Supply that sits at 40% (that’s Short Volume, the supply chain of the stock market), way below market levels.

One would have thought a startup would be the assailant breaching the great Google barrier, but no. It’s the old institution.  Yet my trading decision-support firm, Market Structure EDGE, is a reason why.  EDGE is built on Azure Blazor from Microsoft.  EDGE and ModernIR both run in Azure Cloud.

We’ve gone back to the future.

I don’t know what’ll happen with Artificial Intelligence. But I pointed out here that machines, not humans, were pricing AI (the stock, not the concept).  That’s measurable, like Supply and Demand.

I’m heading somewhere, and I’ve not forgotten about FRC. 

In the past week, I related three things to EDGE users in daily Market Desk notes. One, risk to stocks would be greatest with new options trading, not with old ones lapsing last week.  That’s because what motivated the use of the old ones was already in the past.  So, the stock market’s fall yesterday on Counterpart Tuesday reflects new but unrealized risk.

That’s important information.  It will recalibrate again at May options-resets.

Second, the market has in the past five years always declined when Demand on our ten-point algorithm metering buying and selling by investors and traders exceeds 6.5 in the S&P 500. That happened Apr 6. SPY was $409 (4,090 SPX).

Yes, it held up well.  And we’re coming to a Ford Pinto, a tap on the bumper, and FRC.

But here we are at $406.  You might say, “Okay, but that’s not REALLY down. That’s statistically immaterial.” 

Maybe so. We’ll see how it goes. Broad Sentiment is 5.7, still above the crucial 5.0 fulcrum.  But barring a sudden refreshing market breeze, it will fall from here.

And third, I said the market was fragile. Finely balanced. That FRC could destabilize it with earnings.

Well, that happened.

Volatility has vanished, a product of machines “crossing the spread” to match bids and offers, causing prices to tighten.  But small spreads are susceptible to sudden trouble. Models adjust – artificial intelligence – to smaller spreads stop crossing it if prices suddenly move (this is the hubris of AI, the discipline, too. It won’t adapt well to sudden change, like humans can).

And all it takes is a tap on the bumper. 

FRC shouldn’t destabilize stocks.  It’s not Lehman Brothers or Bear Stearns.  I’m not calling FRC a Ford Pinto. Quite.  But it should be…unsettling…that a small institution can fracture market stability.  Stocks are too dependent on borrowing. Supply. Short Volume.

The same is true in the country, across the planet.  Think about it.

Surreal

Posted on April 5, 2023

 

Sometimes you don’t even know what to say. 

And I’m not talking about politics, geopolitics, the legal system, bills of attainder, the economy, the public fisc, wars, rumors of wars, storms.

Pick your thing.

Photo 82175923 / Artificial Intelligence © Maxuser2 | Dreamstime.com

I mean that during the month of March, according to Vanda Research and reported in the Wall Street Journal, trading by retail investors fell by half versus February.  At the same time (in the same article), flows to Money Market funds neared $200 billion, a record, says the Investment Company Institute.

According to data compiled by Yardeni Research from ICI data, there have been negative flows in US equities every week of 2023.  Negative flows are the same as outflows.

Two major banks have failed and CNN reported Mar 12 that banks are sitting on $620 billion of unrealized balance-sheet losses, the product of not having to mark them to market – unless they have to sell holdings.

Short Volume in S&P 500 stocks has been over 50% of trading volume since Feb 15 and has been below that level at all just 24 trading days in 2023.  Yet somehow the S&P 500 is up 7.4% YTD. The Communications Services and Technology sectors are up 21% each, Consumer Discretionary up 14.5%.

Surreal?

It follows a recurring theme.  It doesn’t matter if fund-flows are up or down in the stock market.  It matters WHERE the money goes.

And it follows a handful of equities.  Just eight big stocks scattered across those three sectors, AAPL, MSFT, AMZN, NVDA, TSLA, GOOG, GOOGL, META, are about 25% of the SPX.

Look how much they’re up.  TSLA and NVDA rose 90%. MSFT is up 20%, AAPL up 32%. AMZN up 20%. You can check the rest.

What’s more, Fast Trading machines comprising over 50% of trading volume in the S&P 500 amplify any move. 

A great example is AI, the company behind ChatGPT.  C3.ai was up 206% YTD before a short attack yesterday dropped it 26%.  But the trading patterns are effectively 100% machines since Mar 16.  Over 56% of daily trading is short – borrowed or created by market-makers.

And it traded 110,000 times per day the past 20 days on average, a number that rose to 170,000 the past five days.  Volatility was 8.6% daily the past month. It trades over $800 million of stock daily, including about $2 billion daily the two days before the short report.

And the whole world has been talking about artificial intelligence and large language models. Some countries have moved to ban it.  There’s a raging debate about the need for rules to address built-in human bias.

AI (the stock, not the acronym) has been the touchstone for how we think about the prospects and probabilities.

And ironically – proving a thing by its opposite – humans haven’t even been trading the stock.  It’s machines using AI. 

It’s amusing, but not for shareholders.

By our behavioral measures, rational people last outcompeted the machines to set price Dec 8. At $12.51. Not once have they done so (they were there but didn’t lead) in 2023.

Which means it was trading 170% higher yesterday than what humans had last indicated they were willing to pay.

That seems pretty important to know, especially if you’re a fiduciary to shareholders like executives and board directors are. 

What thing in your stock’s valuation, good or bad, is a product of AI that doesn’t reflect reality?

Is the market a good barometer for bank risk, economic risk, if money has been leaving it but stocks have been rising, driven by machines, and concentration?

These are the existential questions public companies and investors should, seems to me, be pondering. Rather than whether TSLA is a good deal at these levels or where to hold an Analyst Day in New York. 

As to the economy and the view on it from Jamie Dimon or Barry Sternlicht (people whose views I respect and follow), one thing is certain.  We don’t spend less than we make as a society.  If we did, we wouldn’t need a central bank.

And the party blowing the budget everywhere is Government, in the universal sense.

I heard Joe Manchin say on TV the other day that the law requires the House of Representatives to offer a budget by mid-February and the Senate and House to reconcile differences on it by the end of April.

Then the White House is to submit a budget to Congress by September 30, the end of the government’s fiscal year (why it’s not aligned with Congressional terms, since the House has the power of the purse, is illogical. But I digress).

“I can’t recall the last time any of that happened,” said Sen Manchin.

No wonder things are a mess.  At least the stock market is understandable, measurable. It’ll tell you the truth. But you have to know what it’s saying. 

If you’re confused, holler. We’ll help. We read what machines do.