Three Peaks and Domed House

Core idea

The three peaks and domed house is George Lindsay's 23-point topping formation, first published in 1971 and almost unique in being designed for indices (the Dow Jones Industrials specifically), not individual stocks. The pattern divides into two halves separated by a sharp "separating decline":

1. Three peaks at the start — often flat-looking on monthly data, though irregular on bar charts.
2. A separating decline — a sharp drop that visually cuts the three peaks away from what follows.
3. A domed house — a two-story structure that builds upward: a base (points 10-14), a first-story wall (15), a first-story roof (15-20), a second-story wall, and a dome (the second story plus head-and-shoulders-like apex).
4. A complete collapse — Lindsay claims the entire gain from the domed house always unwinds, returning to roughly the base of the first story.

The pattern is rare in individual stocks. Bulkowski couldn't find any in his first or second edition. He includes it here because it does appear in the major indices, and the 7-month forecast window from the base of the separating decline to the dome peak has been surprisingly accurate in the examples documented.

Why it matters

Most chart patterns target a single stock and a horizon of weeks to months. Three peaks and domed house targets the broader market over a roughly two-year structure. If you can identify one forming in an index, you have a long window to:

• Buy protective puts on long equity positions.
• Sell covered calls against existing longs.
• Build a position in inverse ETFs as the dome completes.
• And critically, stand ready with cash near point 28 (the post-collapse low) for the recovery trade.

It is portfolio-level pattern recognition, not single-stock signal generation.

The 7-month rule

Lindsay's most testable claim is timing: the dome peak (point 23) arrives 7 months and 8-10 days after the base of the separating decline (point 14). When the base isn't well-formed, Lindsay says you can use the earlier valleys (point 6, or even point 4) as the starting clock. Bulkowski tested this on the 2012 Dow chart with point 6 as the anchor and the forecast missed the actual peak by only 11 days.

Key takeaways

Mental model

Copy sourceZoom

Practical application

Searching for the pattern

Example

Suppose the S&P 500 has run from 3,800 to 4,800 over 18 months. Imagine three rough peaks form: 4,750 in March, 4,810 in May, 4,790 in August (the three peaks). A separating decline drops the index to 4,450 in October — below the May-August valley at 4,580. The foundation builds from 4,450 to 4,520 by December.

Apply Lindsay's forecast: 7 months and 8 days from the foundation low (December 1) → around July 9 of the following year.

A defensive playbook:

• April: start scaling into long-dated SPX puts as the rebuilding "first floor" climbs back above 4,700.
• May-June: as the index pushes into the 4,850-4,950 range (the dome), increase put coverage. Sell covered calls against single-stock longs.
• Late June-early July: if the dome peak completes within two weeks of July 9, treat the pattern as confirmed. Move equity allocation to defensive sectors.
• Decline to ~3,800 (= point A): redeploy cash into broad-market longs as price stabilizes near the expected floor.

If the forecast is wrong by 30+ days or the dome never quite forms, the cost is the option premium and some opportunity cost — small relative to the avoided drawdown if the pattern plays out.

Related lessons