the data showing 'staying invested beats timing the market' is mostly run on 1990-2024 windows that conveniently include the bull market structure of post-bretton woods, post-volcker disinflation, and ZIRP. those regimes produced specific returns.
the math gets uglier when:
\- you're investing during regime change (1965-1982 stagflation: inflation-adjusted returns went sideways for 17 years)
\- your time horizon is shorter than the regime cycle (5-10 years vs 15-30 year regime length)
\- the index is at all-time-high concentration (top 10 names = \~35% of S&P)
i'm not saying time the market. i'm saying the rule is closer to 'pick a horizon longer than your regime, and accept the structural sideways risk if you're shorter.' which most retail people don't actually map onto their own situation.
curious if anyone here actually back-tests their assumed return distribution against the 1965-82 window or the 1929-1954 window, or just the post-2009 one. because the answer there changes a lot of allocation defaults