Shifting calendar regimes and election-cycle analogs

A living set of seasonal indicators

The 2004 calendar work treated seasonal indicators as a living set. Some readings were retired, some were newly built, and some older unused readings were revived. The archive presented that upkeep as part of the method, not as a side note after a fixed seasonal proverb.

When cash-flow plumbing moved the peak

A former turn-of-month five-day rise was described as having given way to a midmonth inflow spike after automatic mutual-fund and retirement-plan contributions became routine. The midmonth inflow spike was the updated cash-flow seasonal, not a second proverb kept beside the older turn-of-month pattern.

January as a year-ahead regime month

January was classified as a year-ahead regime month because inaugurations, new Congresses, and policy agendas cluster then. That clustering is the basis of the January year-direction rule. A first-quarter break of the prior December low was treated as a cautionary seasonal signal, the December-low caution, and was compared with that January reading.

The years 2001 and 2003 were named as cases when outside events overrode the January reading, and five major historical errors were acknowledged. The January year-direction rule was therefore a regime clue, not a closed case for the rest of the year.

The stronger six months and the autumn turn

The November-through-April span was identified as the historically stronger six-month stretch, the best six months window, with November, December, and January ranked as the three strongest months and a typical brief pause in February.

By 2004, September was classified as the weakest month and October as a volatile turning month, the October turn month. Since World War II, 10 of the then-recent bear markets were said to have bottomed in October. Weakness and turning were not treated as the same autumn regime.

The four-year election cycle as an overlay

The four-year election cycle was treated as a regime overlay independent of party. The first two years, including the post-election year regime, were associated with wars, recessions, and bear markets, and the latter two years with expansion and bull markets. The years 2001 and 2003 were offered as matching cases, while the 1990s were cited as a technology-driven exception.

After 1952, election years were described as having only one losing final-seven-month stretch, in 2000 during an unresolved contest. October strength was linked to incumbent-win years and November strength to incumbent-loss years. That pairing is the incumbent October-November split.

A long double-bottom analog

The post-bubble 2002-2005 sequence was compared with 1970-1974 and 1929-1932 as a long double-bottom analog. In that comparison, an early low and a later second low, not a single washout, completed the bear regime. The analog was a map of how a long bear could finish, not a claim that one calendar proverb had already ended it.

Read the map with the rest of the market

Seasonal maps were not used in isolation. They were to be read with tape, sentiment, fundamentals, technicals, and monetary policy. Sector seasonality was taken from long-running sector indexes rather than from short exchange-traded-fund histories.