Setup: a trailing-year range ≤ 20% wide held for 150+ sessions without a new 1-year high · trigger = a close to a fresh 252-day high (breakout) · 2000-01-03→2026-07-31
Now: BRK-B at 96% of a 11.9% trailing-year range · 15 months without a new high
Close 511.54 · range 459.11–513.81 · breakout level ≈ 513.81 (-0.4% away) · 312 sessions since the last new 252d high
The finding. A breakout here is not a statistical edge. Pooled across 32 large caps (74 long-range breakouts), the pattern returns +1.2% at 21d and +3.8% at 126d — but that is +0.1 / -2.3 pts vs the stock's own baseline drift. The 73% win rate is large-cap up-drift, not breakout alpha. BRK-B's own 5 breakouts were a coin flip with real failure tails.
Forward returns — breakout vs near-top vs baseline
Median forward return (% positive) at 5 / 10 / 21 / 42 / 63 / 126 sessions. Pooled breakout and pooled near-top are the cross-sectional base rate across 32 large caps; BRK-B breakouts is the name's own history; dashed gold is the unconditional baseline. Watch how the breakout line hugs — and past 42d slips under — the baseline.
Every time BRK-B broke to a new 1-year high out of a mature tight base — the thin sample behind the grey line. Note the two that failed straight into drawdowns.
Breakout date
Range
Base (d)
5d
10d
21d
42d
63d
126d
2006-05-17
13%
528
+2.0%
+1.3%
+1.0%
-0.4%
+4.0%
+17.7%
2014-03-05
17%
155
+5.2%
+3.1%
+4.5%
+6.4%
+6.9%
+15.8%
2016-08-11
19%
413
+0.7%
+0.6%
+0.8%
-2.6%
+2.4%
+11.0%
2018-09-18
20%
157
-0.1%
-1.4%
-3.6%
-0.4%
-10.0%
-6.4%
2019-11-07
18%
271
-1.5%
-2.7%
-0.6%
+2.7%
+2.0%
-21.6%
Read it honestly. For a low-beta compounder like BRK-B, a range breakout is a resumption of the grind, not an explosive move — so it carries no momentum edge, and the failure cases are asymmetric (bull traps that gave back 10–20%). The pooled win rate looks healthy only because large caps drift up regardless. BRK-B's own sample is thin (n=5) and its two failed breakouts ran straight into macro drawdowns. Net: the setup is real and the odds mildly favor up, but the breakout itself is not the edge — if you are long here you are long the drift, not the pattern. Cross-check against the TEF mandate; size for the bull-trap tail, exit-before-entry.
Computed live from adjusted-close history (yfinance). Range width = (252d high − low) / low. Breakout = close ≥ the prior 252-session high after a ≤20% base held 150+ sessions. Pooled base rate spans 32 large caps. Educational research, not investment advice.