Takeaways

#
  • An oversold reading describes a condition; it does not establish that selling has ended.
  • Thompson treated TLT's RSI divergence as incomplete without follow-through.
  • His breadth model required an underlying uptrend and a recovery above the oversold threshold.
  • Weak individual stocks and a resilient capitalization-weighted index can coexist.
  • The September 18 breadth trigger had not yet fired in his model.
Chapters
  1. Define what oversold means
  2. TLT's downtrend and RSI divergence
  3. How the breadth recovery trigger works
  4. Index performance versus individual stocks
  5. Technology as the stronger part of the market

Calling something oversold can sound like an explanation and a recommendation at the same time. Larry Thompson wanted to separate the two.

In his September 18 edition of Thompson's Two Cents, he argued that an oversold reading describes a condition. It does not, on its own, supply the point where that condition ends. Buying solely because an indicator is low can leave an investor without a clear reason to exit if the decline continues.

His preferred question was more demanding: had price or breadth actually begun to recover, and did that recovery create a level against which to manage risk?

The underlying trend changes the signal

Thompson began with the S&P 500's larger uptrend. He thought that context mattered when interpreting weak sentiment and choppy trading beneath the surface. A correction inside an uptrend did not automatically mean the primary trend had ended.

He also objected to leaving the word oversold undefined. If the reference was RSI below 30, say so. If it was a breadth measure, identify the measure and threshold. Without that definition, the term could shift to suit the argument.

Even a defined threshold needed context. A market in a downtrend could produce repeated oversold readings. Their frequency might say more about persistent weakness than about the arrival of a bargain.

Thompson described his own historical work as a reason to be cautious about expecting an immediate bounce. The initial days after an oversold reading could remain difficult. What interested him was the reversion out of that condition, not merely entering it.

TLT showed why patience mattered

Treasuries supplied his example. TLT had generated multiple oversold readings and failed breakouts while remaining in a downtrend. Thompson could see a possible RSI divergence, but he did not think that was enough to enter.

He wanted follow-through: evidence that the bearish momentum regime was changing and price was responding. Until then, a divergence was something to monitor rather than a completed turn.

That distinction protected the idea from becoming circular. If a market is bought because it has fallen, then falls further, it can always be described as an even better bargain. A recovery condition gives the trader something observable to require instead.

A breadth trigger with an exit built in

Thompson explained a stock-market model that first required a favorable larger trend, with the 50-day moving average above the 200-day. Within that context, he tracked the proportion of S&P stocks above their 20-day averages.

His oversold threshold for that breadth measure was 30. The intended trigger came on the recovery through it. "It's that you buy the reversion back above it," he said.

That supplied more than a bullish label. The turn created a nearby reference for a failed recovery, which made the risk easier to define. On September 18, he said the breadth trigger had not yet fired.

An index can hide difficult individual stocks

Thompson also challenged the claim that bearish traders must be wrong whenever the index sits near a high. An investor shorting weak individual stocks can have a different experience from someone shorting the capitalization-weighted index.

His charts showed substantial drawdowns and failed breakouts beneath the surface. At the same time, large technology stocks were holding up well enough to support the indexes. Both observations could be true.

He was particularly interested in technology's consolidation and in whether large names such as Google could resolve higher. Meanwhile, he wanted the weaker sectors to show an actual recovery before treating their declines as buying opportunities.

The approach did not require a fixed bullish or bearish identity. "The market has a nuance to it," Thompson said. His task was to preserve that nuance in the entry: recognize the longer trend, define the condition being measured, and wait for a turn that supplied a usable boundary if it failed.

Questions and answers

#

From the September 18, 2026 conversation.

Why did Thompson wait for a move out of oversold territory?

An oversold reading describes a condition, not its endpoint. Repeated low readings can persist in a downtrend. Thompson wanted price or breadth to recover in a way that created a nearby invalidation level, giving the entry a clearer risk boundary than buying solely because something had fallen.

Link to this answer

How did Thompson's stock-market breadth trigger work?

His model first required the 50-day moving average to be above the 200-day. It then watched the share of S&P stocks above their 20-day averages recover through 30 after an oversold reading. He said that trigger had not yet fired in the September 18 presentation.

Link to this answer

Adapted by Stock Market TV Editorial from the original video. Source date: . Article published: .