How to Trade the AI Panic

Preface: Explaining our market timing models

We maintain several market timing models, each with differing time horizons. The “Ultimate Market Timing Model” is a long-term market timing model based on the research outlined in our post, Building the ultimate market timing model. This model tends to generate only a handful of signals each decade.

 

The Trend Asset Allocation Model is an asset allocation model that applies trend-following principles based on the inputs of global stock and commodity prices. This model has a shorter time horizon and tends to turn over about 4-6 times a year. The performance and full details of a model portfolio based on the out-of-sample signals of the Trend Model can be found here.

 

 

My inner trader uses a trading model, which is a blend of price momentum (is the Trend Model becoming more bullish, or bearish?) and overbought/oversold extremes (don’t buy if the trend is overbought, and vice versa). Subscribers receive real-time alerts of model changes, and a hypothetical trading record of the email alerts is updated weekly here. The hypothetical trading record of the trading model of the real-time alerts that began in March 2016 is shown below.

 

 

The latest signals of each model are as follows:
  • Ultimate market timing model: Buy equities (Last changed from “sell” on 28-Jul-2023)
  • Trend Model signal: Bullish (Last changed from “bearish” on 27-Jun-2025)
  • Trading model: Bullish (Last changed from “neutral” on 18-Nov-2025)

Update schedule: I generally update model readings on my site on weekends. I am also on X/Twitter at @humblestudent and on BlueSky at @humblestudent.bsky.social. Subscribers receive real-time alerts of trading model changes, and a hypothetical trading record of those email alerts is shown here.
 

Subscribers can access the latest signal in real time here.

 

 

A Letdown from the NVIDIA Party

The sense of disappointment was palpable. After the AI market leader reported its earnings, which beat expectations, CEO Jensen Huang characterized demand for its Blackwell chip as “off the charts”. The stock rallied but sellers appeared and pushed the broad market down.

 

A reader characterized Thursday’s market action as, “NVIDIA threw a party but nobody came”. Even before the NVIDIA report, sentiment had been deteriorating. While the Fear & Greed Index is an imperfect sentiment and market timing indicator, its readings have fallen to levels last seen during the “Liberation Day” panic in April. The key difference is fundamentals. The market weakness in April was attributable to fears that global trade might seize up and economic growth may crater. Today’s market weakness is driven mostly by psychology and not a significant change in fundamentals. Should you be stampeding to the exit because some investors are taking profits?

 

I believe sentiment has fallen to levels consistent with a tactical bottom.
 

 

 

Technical Warnings

To be sure, downside risk warnings have been appearing. Breadth deteriorating began in early July and has been warning of possible market weakness for months.
 

 

The NYSE High-Low Logic Index recently spiked, which is a signal of a bifurcated market as both 52-week highs and lows are surging at the same time. Chartists who follow this indicator may be seduced by a case of recency bias. In the last five years, elevated readings have always resolved in corrective market action (pink vertical lines). Go back a further five years, the track record of this indicator is spotty. There were nine signals during that study period, the market resolved five in a benign manner (grey lines) and four in a correction (pink lines).
 

 

The bifurcated nature of market breadth also gave rise to a most refined warning of the Hindenburg Omen, which is triggered when a bifurcated market begins to turn down from an uptrend. A single Omen signal can be ignored, but a cluster of signals, which is what investors saw recently, can warn of corrective action. Similar to the NYSE High-Low Logic Index, Hindenburg Omens suffer from a recency bias problem in which recent signals saw pullbacks (pink) while a longer-term history during the 2015–2019 period saw false positives (grey lines).
 

 

The market has sustained technical damage to the bull trend when the S&P 500 violated its 50 dma support. Even a rally here could form a possible head and shoulders pattern.
 

 

 

Tactical Buy Signals

Nevertheless, I are seeing tactical buy signals everywhere. When the S&P 500 reversed lower after the initial surge Thursday after the NVIDIA earnings report, volume in SPY, the S&P 500, surged to levels consistent with short-term panic bottoms. At a minimum, past episodes have seen short-term bullish reversals.
 

 

Thursday’s market was highly unusual inasmuch as the S&P 500 gaped up over 1% but closed red on the day. Though the sample size is small (n=7), historical studies of similar events have seen bullish resolutions on a time frame of a week or more.
 

 

Three of the five components in my Bottom Spotting Model have triggered. Historically, traders have seen long entry points with positive risk/reward ratios whenever two or more components flashed buy signals.
 

The VIX Index spiked above its upper Bollinger Band, indicating an oversold market, and the NYSE McClellan Oscillator (NYMO) fell to an oversold level. As well, the term structure of the VIX Index has inverted, indicating fear.
 

 

As well, the Zweig Breadth Thrust Indicator fell to an oversold reading last week. While ZBT buy signals are rare, as they require market breadth to surge from oversold to overbought within 10 trading days, oversold ZBT Indicator signals are not. Oversold readings in the last two years have seen the market bounce every single time.
 

 

As well, the reaction to Thursday’s market action was indicative of capitulations that mark short-term bottoms, judging by the number of discussions I had with readers on that day.
 

 

Fundamental and Macro Support

From a fundamental perspective, the market price retreat afforded the stock prices some valuation support. The S&P 500 forward P/E ratio came off the boil and fell to 21.5, though it’s still above its 5- and 10-year averages.
 

 

The decline in P/E was further supported by an increase in earnings estimates, indicating positive fundamental momentum.
 

 

From a top-down macro perspective, the liquidity environment is showing some signs of improvement, which should be positive for risk assets. Reduced expenditure from the U.S. government shutdown translated to a rising Treasury General Account. Now that the shutdown is over the injection of TGA cash into the economy will boost liquidity.
 

 

As well, the New York Fed’s Reserve Demand Elasticity has been surging, which is a sign of bank reserves are becoming less abundant. The Fed has become concerned and it is taking a number of steps, such as ending quantitative tightening, to address the problem. If it persists, watch for money market intervention designed to boost financial system liquidity.
 

 

Bitcoin is a sensitive indicator of liquidity. From a technical perspective, it has fallen to a support zone, and the Daily Sentiment Index (DSI) is oversold with a reading of 13.
 

 

In conclusion, the recent market pullback may have been attributable to a combination of breadth deterioration and a highly bifurcated market. In the short term, technical price action and sentiment have become stretched to the downside that a bounce is more or less inevitable. I continue to believe stock prices will rally into year-end, but I am watching for signs of a bullish follow-through after the reflex rally for confirmation.

 

 

Disclosure: Long SPXL
 

3 thoughts on “How to Trade the AI Panic

  1. The forward P/E chart was repeated twice in the original post. The second chart has been replaced with the correct one.

    I apologize for the error.

  2. Cam-
    The last line of your update is confusing.
    “The trading model is not be neutral.” Is this is a typo?

Comments are closed.