Reactions to Reactions
to Reactions

By Scott Covert · independent analyst & builder of the AI Stock Market Impacts engine · Ontario, Canada

Why everyone predicting the market at the same time makes short-term forecasting impossible. And why that's actually good news for us.

Here's the dirty secret of short-term stock forecasting:

Everyone is trying to predict everyone else's predictions.

Traders react to other traders' reactions. Algorithms front-run algorithms. Analysts revise forecasts based on how other analysts revised their forecasts. It's a confounding spiral where the signal drowns in its own echo.

The market moves up and down. Of course it does. But in the short term, those moves aren't driven by what companies are actually doing. They're driven by what millions of people think millions of other people are about to do.

That's not forecasting. That's a hall of mirrors.

7 Reasons Short-Term Forecasting Is Broken

1

The Reflexivity Spiral

Everyone's predicting everyone else's predictions. Trades are reactions to reactions to reactions. By the time you act on "the signal," you're trading on the echo of an echo.

2

The Algorithmic Arms Race

High-frequency firms spend billions shaving microseconds. Retail investors and even many institutional players are playing a game where the rules change faster than they can read them. You're bringing a knife to a laser fight.

3

News-Driven Chaos

A tweet, an earnings whisper, a Fed chair's tone of voice can move markets 3% in minutes. No model reliably predicts human impulsiveness. And no model ever will.

4

Herding and Panic Cycles

Short-term markets are driven by fear and greed, not fundamentals. Behavioral contagion makes crowds irrational in ways that are mathematically unpredictable. Monday's genius is Tuesday's bagholder.

5

The Overfitting Epidemic

Every quant fund's backtested model looks brilliant... until it doesn't. Short-term patterns are mostly noise dressed up as signal. If backtesting worked, every hedge fund would be up 40% a year. They're not.

6

Information Asymmetry Decay

By the time retail investors act on "news," institutional money has already priced it in. You're always trading on yesterday's insight, convinced it's today's edge.

7

Liquidity Illusions

Markets look liquid until everyone runs for the exit at once. Short-term models assume orderly markets that don't exist during the only moments that actually matter.

THE TIME HORIZON FLIP

Short-Term vs Long-Term

8 Reasons 3–5 Year Forecasting Is Easier

1

Macro Trends Are Visible and Slow-Moving

AI adoption curves, industry digitization, demographic shifts, energy transitions — these don't reverse on a Tuesday. They compound predictably over years. The direction of the river is knowable, even when the rapids aren't.

2

Mean Reversion Works at Scale

Short-term noise washes out. Over 3–5 years, industries gravitate toward where the real productivity and value creation is happening. The signal emerges from the noise.

3

Technology Adoption Follows Known S-Curves

We have 200+ years of data on how transformative tech propagates through industries. The pattern — early adopters, acceleration, saturation — is remarkably consistent from the steam engine to smartphones to AI.

4

Capital Expenditure Is Committed

When companies pour billions into AI infrastructure, that money is spent. The effects play out over years regardless of quarterly sentiment swings. You can't un-build a data center because the NASDAQ dipped.

5

Competitive Dynamics Clarify

Winners and losers in an industry disruption become obvious at the 3–5 year horizon. The fog of "who will win AI?" clears as deployment data accumulates. Today's bets, tomorrow's certainties.

6

Regulatory Frameworks Crystallize

Policy responses to new tech are chaotic in year 1, but by year 3–5 the rules are largely set. You can model around known constraints instead of guessing at unknown ones.

7

Cross-Industry Cascading Becomes Measurable

When AI transforms logistics, the downstream effect on retail, manufacturing, and agriculture takes 2–4 years to show up in financials. Which means it's forecastable before it's priced in. That's the window we operate in.

8

The Confounding Spiral Weakens

Day traders aren't competing with you at the 5-year horizon. Most market participants are playing a completely different, shorter game. Less competition for the same insight means more signal, less noise.

The irony: the time horizon everyone ignores — because it doesn't feel urgent, because it doesn't trigger dopamine, because it doesn't look like "trading" — is the one where forecasting actually works.

Short-term prediction is a war of reflexes.
Long-term prediction is a war of understanding.

That's where deep AI industry analysis has a real edge. Not predicting tomorrow's close. Predicting which industries will be structurally transformed over the next 3–5 years, and positioning accordingly. 167 cross-industry effects. 28 industries. 8 analytical dimensions. Updated continuously.

What this looks like in real numbers

The smartest forecasters on Wall Street, calling 2023

This isn't a theory you have to take on faith. It plays out, on the record, every single year. Take the entry into 2023 — a year when the people paid the most to forecast the market lined up and pointed the same direction, and the market walked off in the other.

Heading into 2023, the consensus from Wall Street equity strategists clustered around a small gain. A Bloomberg survey of strategists put the average year-end S&P 500 target near 4,009; a Reuters poll of 41 strategists landed at a 4,200 median. Both implied roughly a high-single-digit move — the same modest, comfortable number these targets cluster around almost every year. The two most-watched bears went further:

Forecaster 2023 year-end S&P 500 call Implied
BNP Paribas — Greg Boutle 3,400 (lowest on the Street) ~11% drop
Morgan Stanley — Mike Wilson 3,900 roughly flat
Strategist consensus (Bloomberg avg.) ~4,009 high single digits
What actually happened 4,769.83 close (Dec 29, 2023) +24% price, ~27% total return

BNP Paribas's Greg Boutle had the lowest target on the Street at 3,400 — and didn't just trim it, he abandoned it mid-year, flipping to 4,150 by September as the rally ran away from him. Morgan Stanley's Mike Wilson, the strategist everyone quoted, held 3,900 deep into the fall. The index closed the year at 4,769.83. Wilson's call missed by more than 800 points. The consensus missed the actual return by roughly 17 percentage points.

And 2023 wasn't the freak year. It was the norm.

Across the strategist community, year-end S&P 500 targets have underestimated the actual outcome in 13 of the past 16 years — missing by about 10 percentage points on average. The number that gets published in December is reliably wrong by the following December, almost always to the downside, because it's anchored to a polite consensus rather than to anything the market is obligated to obey.

The economists fared no better. In October 2022, Bloomberg Economics' own recession model — thirteen macro and financial indicators, run by economists Anna Wong and Eliza Winger — put the probability of a U.S. recession within twelve months at 100%. Not 80, not 95. One hundred. There was no recession. The economy grew, unemployment stayed low, and by late 2023 the same forecasters were quietly writing up why everyone got it wrong.

Look at what every one of these misses has in common. Each forecaster was reading the same Fed signals, the same inflation prints, the same yield curve — and reading each other. A 3,400 target gives the next strategist permission to print 3,600; a 100% recession model gives the next desk cover to stay bearish. That's the hall of mirrors from the top of this page, with brand names and dollar figures attached. At a twelve-month horizon, the dominant input isn't the economy. It's the other forecasts — reactions to reactions, priced in before the ink dries.

So the engine on this site doesn't try to win that game. It deliberately skips the month-to-month price call, where reflexivity rules and even the best-resourced desks miss by double digits, and works the 3–5 year structural layer instead — committed capital, adoption curves, cross-industry cascades that take years to show up in financials. Not because long horizons are easy, but because they're the one place the noise finally has time to wash out.

Sources: S&P 500 strategist targets and 2023 dispersion — Financial Samurai, Bloomberg (BNP Paribas), CNBC (Morgan Stanley), CNBC (2023 close). Bloomberg Economics 100% recession model — Bloomberg, with the post-mortem in Bloomberg Opinion.
Limited Access

We're opening the full engine to 50 testers.

Free. You test it, explore every industry, run scenarios. We learn from how you use it.

28 industries scored across 8 analytical dimensions. 167 cross-industry cascade effects. Custom scenario modeling. The same engine we charge $279/year for — yours free, limited to 50 people.

Get Free Access Takes 30 seconds. No credit card.

Not ready to test? Run a free portfolio scan — enter your tickers and see how your industries score.

About the author

I'm Scott Covert — an independently curious person and the person who built everything here, including the 28-industry cross-effect engine — the “AI Revolution Cascade Matrix”. I'm not a fund, a broker, or a newsletter reselling someone else's research. I built the systems that take my ideas and sources and turn them into opinion pieces with machine-verified reasoning and sources, all shown so you can argue with me (I am, after all, trying to predict the future of the stock market, through a series of continual deep research loops into everything affecting stocks).

My edge is pattern recognition across fields (an involuntary feature of ADHD), not a Wall Street pedigree. Everything here is directional synthesis meant to help you think, not financial advice. (If you're a publication or fund and want to license or collaborate, that lives over here.)